url
stringlengths 52
53
| repository_url
stringclasses 1
value | labels_url
stringlengths 66
67
| comments_url
stringlengths 61
62
| events_url
stringlengths 59
60
| html_url
stringlengths 40
43
| id
int64 719M
2.8B
| node_id
stringlengths 18
32
| number
int64 1.28k
11.9k
| title
stringlengths 1
350
| user
dict | labels
listlengths 0
10
| state
stringclasses 2
values | locked
bool 2
classes | assignee
dict | assignees
listlengths 0
3
| milestone
null | comments
sequencelengths 0
30
⌀ | created_at
timestamp[s] | updated_at
timestamp[s] | closed_at
timestamp[s] | author_association
stringclasses 4
values | sub_issues_summary
dict | active_lock_reason
stringclasses 1
value | draft
bool 2
classes | pull_request
dict | body
stringlengths 0
73.4k
⌀ | closed_by
dict | reactions
dict | timeline_url
stringlengths 61
62
| performed_via_github_app
null | state_reason
stringclasses 3
values | is_pull_request
bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/NVIDIA/NeMo/issues/1475 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1475/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1475/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1475/events | https://github.com/NVIDIA/NeMo/issues/1475 | 746,992,334 | MDU6SXNzdWU3NDY5OTIzMzQ= | 1,475 | [Question] example of ConformerEncoder class | {
"login": "Omarnabk",
"id": 72882909,
"node_id": "MDQ6VXNlcjcyODgyOTA5",
"avatar_url": "https://avatars.githubusercontent.com/u/72882909?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Omarnabk",
"html_url": "https://github.com/Omarnabk",
"followers_url": "https://api.github.com/users/Omarnabk/followers",
"following_url": "https://api.github.com/users/Omarnabk/following{/other_user}",
"gists_url": "https://api.github.com/users/Omarnabk/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Omarnabk/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Omarnabk/subscriptions",
"organizations_url": "https://api.github.com/users/Omarnabk/orgs",
"repos_url": "https://api.github.com/users/Omarnabk/repos",
"events_url": "https://api.github.com/users/Omarnabk/events{/privacy}",
"received_events_url": "https://api.github.com/users/Omarnabk/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"You may use this example to use the Conformer encoder with CTC-loss:\r\nhttps://github.com/NVIDIA/NeMo/blob/main/examples/asr/experimental/conformer/conformer_linear_decoder_bpe.yaml\r\n\r\nWe have the plan to release some checkpoints trained with Confomer-CTC in our next release in March, maybe even Conformer-RNNT."
] | 2020-11-19T23:01:30 | 2021-05-28T08:45:06 | 2021-05-28T08:45:06 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Are you going to provide examples on how to use the newly added ASR decoders, like ConformerEncoder? | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1475/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1475/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1474 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1474/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1474/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1474/events | https://github.com/NVIDIA/NeMo/pull/1474 | 746,964,828 | MDExOlB1bGxSZXF1ZXN0NTI0MjY1NTg3 | 1,474 | Added storage of intent and slot labels inside of the config and did … | {
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@vladgets and/or @tkornuta-nvidia could you please resolve merge conflicts?",
"Hey @okuchaiev I am not responsible for this PR.\r\n\r\nAll the TLT-related changes to this model that I need were incorporated in #1477 .\r\n\r\n@vladgets it is your shot on what you want to do with this.",
"Since Tomasz did his changes in the code, I will just cancel this PR request."
] | 2020-11-19T22:10:28 | 2020-12-02T00:07:33 | 2020-12-02T00:07:22 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1474",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1474",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1474.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1474.patch",
"merged_at": null
} | Added storage of intent and slot labels inside of the config and did correction in prediction function
| {
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1474/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1474/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1473 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1473/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1473/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1473/events | https://github.com/NVIDIA/NeMo/pull/1473 | 746,909,988 | MDExOlB1bGxSZXF1ZXN0NTI0MjIwNTQw | 1,473 | Improve RNNT Greedy decoding | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-19T20:42:49 | 2020-11-21T09:11:06 | 2020-11-21T09:11:01 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1473",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1473",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1473.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1473.patch",
"merged_at": "2020-11-21T09:11:01"
} | # Changelog
- As discussed in https://github.com/NVIDIA/NeMo/issues/1458, greedy decoding can be simplified by reusing `blank_mask` rather than maintaining both `blank_mask` and `time_mask`.
Closes https://github.com/NVIDIA/NeMo/issues/1458 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1473/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1473/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1472 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1472/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1472/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1472/events | https://github.com/NVIDIA/NeMo/pull/1472 | 746,908,013 | MDExOlB1bGxSZXF1ZXN0NTI0MjE4OTk0 | 1,472 | Enable `default` pytorch initialization mode for ConvASREncoder/ConvASRDecoder modules | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@titu1994, please merge when ready\r\n"
] | 2020-11-19T20:39:37 | 2020-11-20T19:51:23 | 2020-11-20T19:17:29 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1472",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1472",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1472.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1472.patch",
"merged_at": "2020-11-20T19:17:29"
} | # Changelog
- Adds `init_mode=None` option which will preserve the default pytorch initialization for the `ConvASREncoder` and `ConvASRDecoder` modules
Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1472/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1472/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1471 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1471/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1471/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1471/events | https://github.com/NVIDIA/NeMo/pull/1471 | 746,859,993 | MDExOlB1bGxSZXF1ZXN0NTI0MTc5Mzcy | 1,471 | Encoder-Decoder NMT Refactor | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 4 alerts** when merging 17d0dd296bac97e6824925ddabcacee075b7fe8b into cfc84cebd98d92fa06089cdd42da685e60521d26 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-d88619b7202d9dc56b09659798de28b51248b54a)\n\n**new alerts:**\n\n* 4 for Unused import",
"This pull request **introduces 8 alerts** when merging d858915a912e78f218bab6cc94c4ffd7f26e8942 into cfc84cebd98d92fa06089cdd42da685e60521d26 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-fac0620f18e4c7816b09ac9699617a22d3891dc8)\n\n**new alerts:**\n\n* 8 for Unused import",
"This pull request **introduces 15 alerts** when merging 25e710ba74a4cf3fe8de1ee91e4da5692f15322d into cfc84cebd98d92fa06089cdd42da685e60521d26 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-cb1027ef7da23281742250f5ec617c752b3e94b7)\n\n**new alerts:**\n\n* 14 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 17 alerts** when merging 9b8c49661679715d867d3f9d3dbd1d66f8164019 into ddfae2e8988cd4d518e241daad1a75660735f2ba - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-e37798307f4f0826ba5748ae1440065385f20778)\n\n**new alerts:**\n\n* 16 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 17 alerts** when merging b0a4ad0d051cfcece9f0093025a9d80c336b718f into ddfae2e8988cd4d518e241daad1a75660735f2ba - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-7cad3cd5bf38ad1c85e4352af7bf04e08d90dede)\n\n**new alerts:**\n\n* 16 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 17 alerts** when merging 6f5888bd158b94ea2701f4f6292bdd81ed4479ca into ddfae2e8988cd4d518e241daad1a75660735f2ba - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-d5afc5df3bbd364e9ef00b343147958b04770cd2)\n\n**new alerts:**\n\n* 16 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 20 alerts** when merging 8cfa7020797a2ccfde516e79baba2d98fa346419 into ddfae2e8988cd4d518e241daad1a75660735f2ba - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-b01d9f00d0e8fd6c8f44a2004610d745d27ea7c7)\n\n**new alerts:**\n\n* 17 for Unused import\n* 3 for Unused local variable",
"This pull request **introduces 21 alerts** when merging 35bd74eefd05d4c3285447e766dc9a043de1d24d into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8076f782ed762815182fd9292e00ebf374b410d8)\n\n**new alerts:**\n\n* 20 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 21 alerts** when merging 24752737851d1897dde01e83367358aaad13f201 into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-979784d5ef428b35c4cb2cc901b29e2105339368)\n\n**new alerts:**\n\n* 20 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 21 alerts** when merging bac2a86c8be4f0251c22957a8ef33ed4e3e628ab into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3ac6b004473f9efcd8ddec8353d3b80b2a967b81)\n\n**new alerts:**\n\n* 20 for Unused import\n* 1 for Unused local variable",
"This pull request **introduces 23 alerts** when merging 8ccd8187fbf801e26e8ade3c2a64676dc3e13197 into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-c1d536eda4b30d51018ecc4a725c606e46c3510d)\n\n**new alerts:**\n\n* 20 for Unused import\n* 2 for Wrong name for an argument in a call\n* 1 for Unused local variable",
"This pull request **introduces 20 alerts** when merging 6a0437917fbf0a08a763281d4875f4ef15975ca7 into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3b152ab26ffa3cba325d89ef133b2b87dc604a83)\n\n**new alerts:**\n\n* 20 for Unused import",
"This pull request **introduces 20 alerts** when merging 98a5ae1e06d88733e57f9e5422091d7e5371c007 into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-e5913671bedb07fc4205107d78658d964ecd6243)\n\n**new alerts:**\n\n* 20 for Unused import",
"This pull request **introduces 14 alerts** when merging 9b4f476a33127d676daffd1d671ac27c5505b55b into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-6f66a05ee92cb7395f1a5624e5b6ef4de1c82936)\n\n**new alerts:**\n\n* 14 for Unused import",
"This pull request **introduces 14 alerts** when merging 5eb4d8817cbc13d694f4ecce94dd1a8f627b0b9a into 631eaf9845d77708d6922dae5b5ac17805d19026 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-66d5ffd00bdaeb24916cc8a46cabc8abedb36627)\n\n**new alerts:**\n\n* 14 for Unused import",
"This pull request **introduces 16 alerts** when merging 0625b5fc64bec78371876ca975fd1446c8fc1e1b into 96ccc9dd0effd784ba3d31298b14785f8e9a93db - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-9b31984501b38705e04a63a26d5dfe966747b95a)\n\n**new alerts:**\n\n* 16 for Unused import",
"This pull request **introduces 16 alerts** when merging cc591b64ef98835bdbba149d2a8a8ed5f7aadeba into 96ccc9dd0effd784ba3d31298b14785f8e9a93db - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-be5cbcb7239b94c1dcad7c31b4990bf7c799a773)\n\n**new alerts:**\n\n* 16 for Unused import",
"This pull request **introduces 19 alerts** when merging 4a4a662aaba4b5357c59e44d07b998a82b90fbc9 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3d5aad3b34e2c27949cc7281b3508552fd29b907)\n\n**new alerts:**\n\n* 19 for Unused import",
"This pull request **introduces 19 alerts** when merging 750d615c8c8c905864016c6acfd326d1b2da97b4 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-f0d280028b2cf66308279a76e1afe843d7683a3d)\n\n**new alerts:**\n\n* 19 for Unused import",
"This pull request **introduces 19 alerts** when merging 37df969755128a7190dcac724e8150772176f658 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-952910ee1718de6add68587c6793c576669e19b5)\n\n**new alerts:**\n\n* 19 for Unused import",
"This pull request **introduces 18 alerts** when merging e81229acad79f89d1918598a9a0c9263b5bc0851 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-596f499013fbd64035308cccf7788d72977a3c25)\n\n**new alerts:**\n\n* 18 for Unused import",
"This pull request **introduces 1 alert** when merging d1b56f60e3f42fc0dedfa57cb0aa437be6d3a6f4 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-89439b9c71975e32c138c2be0ae9ff20fbf1b95a)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 12 alerts** when merging 8111e38ea99ffcd056e5c2c8f1f377adf7c22353 into cf6d49c7d84b2edb2c48e9ef09fa088f8379a0c3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-5dd19041e5d3723dec2ae64bbe16fd75f64bac1b)\n\n**new alerts:**\n\n* 12 for Unused import",
"This pull request **introduces 5 alerts** when merging 08966b703988212246ef3943f5acf09bedd4983e into c1330f384a9d4a986cc48ad94f079e39190fd49b - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-d6bb2c644884ed551798475f6a4e284c52402cb7)\n\n**new alerts:**\n\n* 5 for Unused import",
"This pull request **introduces 2 alerts** when merging 1711319c67ad3513f45df07b45cdaa6095d8a0df into c1330f384a9d4a986cc48ad94f079e39190fd49b - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-61d27e0f98acf155410b3827bf82564e7a683297)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging 88021aea642061ed9c16d7d2771edd82b7ac6598 into c1330f384a9d4a986cc48ad94f079e39190fd49b - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-4dcdf0277a67d6b2f8f33d47f8432d003c03164c)\n\n**new alerts:**\n\n* 2 for Unused import"
] | 2020-11-19T19:26:41 | 2020-12-04T06:34:58 | 2020-12-04T06:34:52 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1471",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1471",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1471.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1471.patch",
"merged_at": "2020-12-04T06:34:52"
} | This PR will refactor the `TransformerMTModel` so that pre-trained HF and Megatron encoders and decoders can easily be used along with the `TransformerEncoder` and `TransformerDecoder` for NMT.
# Plan for this PR
- rnd -> rnd
# Plan for future PRs
- pretrained encoder -> rnd
- pretrained encoder -> pretrained decoder
# TODO
- [ ] Organize machine translation naming and folder structure
- [x] Jenkins test
- [ ] Unit Tests
- [ ] Instantiate tokenizers with hydra
# Usage
```
python examples/nlp/machine_translation/enc_dec_nmt.py \
--config-path=${CONFIG_PATH} \
--config-name=${CONFIG_NAME} \
model.max_generation_delta=10 \
model.train_ds.src_file_name=/raid/data/68792/train.clean.en.shuffled.dev \
model.train_ds.tgt_file_name=/raid/data/68792/train.clean.de.shuffled.dev \
model.train_ds.tokens_in_batch=8192 \
model.validation_ds.src_file_name=/raid/data/68792/wmt14-en-de.src \
model.validation_ds.tgt_file_name=/raid/data/68792/wmt14-en-de.ref \
model.validation_ds.tokens_in_batch=1024 \
model.test_ds.src_file_name=/raid/data/68792/wmt14-en-de.src \
model.test_ds.tgt_file_name=/raid/data/68792/wmt14-en-de.ref \
model.encoder_tokenizer.tokenizer_model=/raid/data/68792/tokenizer.BPE.37K.model \
model.decoder_tokenizer.tokenizer_model=/raid/data/68792/tokenizer.BPE.37K.model \
trainer.gpus=4 \
trainer.val_check_interval=1.0 \
``` | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1471/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1471/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1470 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1470/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1470/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1470/events | https://github.com/NVIDIA/NeMo/pull/1470 | 746,772,672 | MDExOlB1bGxSZXF1ZXN0NTI0MTA3NTI3 | 1,470 | updates for test mode for TLT | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-19T17:14:50 | 2020-11-20T21:01:16 | 2020-11-20T21:01:12 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1470",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1470",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1470.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1470.patch",
"merged_at": "2020-11-20T21:01:12"
} | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1470/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1470/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1469 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1469/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1469/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1469/events | https://github.com/NVIDIA/NeMo/pull/1469 | 746,161,981 | MDExOlB1bGxSZXF1ZXN0NTIzNjA3ODQ0 | 1,469 | Add character coverage to SPE for ASR/NLP custom tokenizers | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-19T01:09:09 | 2020-11-19T02:31:10 | 2020-11-19T02:31:02 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1469",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1469",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1469.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1469.patch",
"merged_at": "2020-11-19T02:31:02"
} | # Bugfix
- Enables sentencepiece tokenizers to capture the entire vocabulary for small vocab languages like English with cases.
Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1469/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1469/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1468 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1468/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1468/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1468/events | https://github.com/NVIDIA/NeMo/pull/1468 | 746,103,800 | MDExOlB1bGxSZXF1ZXN0NTIzNTYwMDc4 | 1,468 | Fix sentencepiece formatting error | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-18T22:59:28 | 2020-11-19T00:08:28 | 2020-11-18T23:49:43 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1468",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1468",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1468.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1468.patch",
"merged_at": "2020-11-18T23:49:43"
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1468/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1468/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1467 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1467/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1467/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1467/events | https://github.com/NVIDIA/NeMo/issues/1467 | 745,618,724 | MDU6SXNzdWU3NDU2MTg3MjQ= | 1,467 | AudioText leaks memory for large-scale training datasets | {
"login": "vadimkantorov",
"id": 1041752,
"node_id": "MDQ6VXNlcjEwNDE3NTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1041752?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vadimkantorov",
"html_url": "https://github.com/vadimkantorov",
"followers_url": "https://api.github.com/users/vadimkantorov/followers",
"following_url": "https://api.github.com/users/vadimkantorov/following{/other_user}",
"gists_url": "https://api.github.com/users/vadimkantorov/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vadimkantorov/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vadimkantorov/subscriptions",
"organizations_url": "https://api.github.com/users/vadimkantorov/orgs",
"repos_url": "https://api.github.com/users/vadimkantorov/repos",
"events_url": "https://api.github.com/users/vadimkantorov/events{/privacy}",
"received_events_url": "https://api.github.com/users/vadimkantorov/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1485815557,
"node_id": "MDU6TGFiZWwxNDg1ODE1NTU3",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": "Something isn't working"
}
] | closed | false | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"Thanks, @vadimkantorov for reporting this. Will look into the solution and test it out. ",
"Indeed, when @squire-tomsk investigated stubbing out `.data` in the UserList to our share-safe list holder in https://gist.github.com/vadimkantorov/86c3a46bf25bed3ad45d043ae86fff57, the memory leak goes away. The other solution with Manager'd ShareableDict objects may require Python 3.8, but it probably should be fine.\r\n\r\nAlso if supporting the full UserList interface is not necessary, it would be best to get rid of it, since slicing would require special care with these shareable list hacks.",
"The leak will be even worse when persistent workers are used",
"@titu1994 Is the issue fixed? Or superseded by another issue? Could you please provide any clarifying information about issue closing? Thank you!",
"@vadimkantorov were you able to solve the issue using your tensorbacked arrays? I tried replacing audio_files in the __init__ function with a StringArray, but it crashes due to high cpu and memory load, probably when iterating through it",
"Another option is to switch to [tarred datasets](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/stable/asr/datasets.html#tarred-datasets). ",
"@samehraban btw do you know if TarDatasets from webdataset/in nemo support caching an index file? so that one does not need to always do a linear scan to discover what's storred in the tarball ",
"@vadimkantorov Do you mean a single cache for all the tar files? I don't think so.",
"A single index for all tar files, or even an index-per-tarfile... For if we have offset+count index, individual files can be read from tarfile by just doing seek"
] | 2020-11-18T12:30:42 | 2023-05-17T22:06:01 | 2021-07-13T15:20:39 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | If I understand well, it appends and stores paths to audio files in a list:
https://github.com/NVIDIA/NeMo/blob/8d506d5e47367643328375e7e5b0ff6643ddb82b/nemo/collections/asr/parts/collections.py#L90
If this object is then scattered to all DataLoader workers, it leads to leaking Linux shared memory (measured by PSS) and gets some data loader threads killed by OOM killer halting the training process. Full explanation of the bug is in: https://github.com/pytorch/pytorch/issues/13246.
Large-scale training datasets contain a lot of audio files and hence a lot of audio path strings, this exacerbates the problem
One solution is to use Manager objects as proposed by @snakers4 in https://github.com/pytorch/pytorch/issues/13246#issuecomment-725488844
Another solution is to pack strings into tensors as done in my gist: https://gist.github.com/vadimkantorov/86c3a46bf25bed3ad45d043ae86fff57#file-tensorbackeddictarray-py-L32 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1467/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1467/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1466 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1466/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1466/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1466/events | https://github.com/NVIDIA/NeMo/pull/1466 | 745,443,077 | MDExOlB1bGxSZXF1ZXN0NTIzMDA3NTM3 | 1,466 | fix for scripts for audio->manifest in SD/VAD | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-18T08:22:56 | 2020-11-18T08:41:08 | 2020-11-18T08:40:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1466",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1466",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1466.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1466.patch",
"merged_at": "2020-11-18T08:40:58"
} | tiny fix write_long_audio_manifest.py
Results are same actually, Just the previous code casue negative offsets in manifests which is not good.
Signed-off-by: fayejf <[email protected]> | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1466/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1466/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1465 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1465/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1465/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1465/events | https://github.com/NVIDIA/NeMo/issues/1465 | 745,341,881 | MDU6SXNzdWU3NDUzNDE4ODE= | 1,465 | [Question] why is _test_dl defiend in setup_training_data() and setup_validation_data() at qa_model.py | {
"login": "alberto-hong",
"id": 26590050,
"node_id": "MDQ6VXNlcjI2NTkwMDUw",
"avatar_url": "https://avatars.githubusercontent.com/u/26590050?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/alberto-hong",
"html_url": "https://github.com/alberto-hong",
"followers_url": "https://api.github.com/users/alberto-hong/followers",
"following_url": "https://api.github.com/users/alberto-hong/following{/other_user}",
"gists_url": "https://api.github.com/users/alberto-hong/gists{/gist_id}",
"starred_url": "https://api.github.com/users/alberto-hong/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/alberto-hong/subscriptions",
"organizations_url": "https://api.github.com/users/alberto-hong/orgs",
"repos_url": "https://api.github.com/users/alberto-hong/repos",
"events_url": "https://api.github.com/users/alberto-hong/events{/privacy}",
"received_events_url": "https://api.github.com/users/alberto-hong/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-18T05:24:10 | 2021-08-17T00:57:21 | 2021-08-17T00:57:21 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | https://github.com/NVIDIA/NeMo/blob/e82ff1863bf750673400592a8220d63dff51933a/nemo/collections/nlp/models/question_answering/qa_model.py#L282
**Describe your question**
NeMo version is '1.0.0b2'
This is a question about setup_training_data() and setup_validation_data() nemo/collections/nlp/models/question_answering/qa_model.py
I wonder why test data loader is set as None if Dataloader config or file_path for the training or validation is missing.
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1465/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1465/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1464 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1464/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1464/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1464/events | https://github.com/NVIDIA/NeMo/pull/1464 | 745,093,262 | MDExOlB1bGxSZXF1ZXN0NTIyNzExMjk3 | 1,464 | [Bug] Ensure lr scheduler uses limit_train_batches when working out max steps | {
"login": "SeanNaren",
"id": 6707363,
"node_id": "MDQ6VXNlcjY3MDczNjM=",
"avatar_url": "https://avatars.githubusercontent.com/u/6707363?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SeanNaren",
"html_url": "https://github.com/SeanNaren",
"followers_url": "https://api.github.com/users/SeanNaren/followers",
"following_url": "https://api.github.com/users/SeanNaren/following{/other_user}",
"gists_url": "https://api.github.com/users/SeanNaren/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SeanNaren/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SeanNaren/subscriptions",
"organizations_url": "https://api.github.com/users/SeanNaren/orgs",
"repos_url": "https://api.github.com/users/SeanNaren/repos",
"events_url": "https://api.github.com/users/SeanNaren/events{/privacy}",
"received_events_url": "https://api.github.com/users/SeanNaren/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Hi, thanks for the bugfix! Could you fix the style issues by running `python setup.py style --scope=nemo/ --fix` ? Oh theres also a style issue in tests so another pass of `python setup.py style --scope=tests/ --fix` ",
"@SeanNaren Thanks for the test fix, youll need to run style fixes as mentioned above to pass the tests.",
"Thanks @titu1994! Should be fixed :) ",
"Perfect, thanks !",
"Thanks @SeanNaren!",
"Thank you guys :)"
] | 2020-11-17T21:15:49 | 2020-11-17T22:27:39 | 2020-11-17T22:27:06 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1464",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1464",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1464.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1464.patch",
"merged_at": "2020-11-17T22:27:06"
} | When `limit_train_batches` is different to the default, this is not included in the inferred calculation of max_steps. This leads to the incorrect schedule being used when `limit_train_batches` is used.
This PR adds `limit_train_batches` to the calculation to ensure the schedule is correct! | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1464/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1464/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1463 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1463/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1463/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1463/events | https://github.com/NVIDIA/NeMo/issues/1463 | 745,032,728 | MDU6SXNzdWU3NDUwMzI3Mjg= | 1,463 | Confidence measures for Quartznet predictions | {
"login": "jonaskratochvil",
"id": 34024373,
"node_id": "MDQ6VXNlcjM0MDI0Mzcz",
"avatar_url": "https://avatars.githubusercontent.com/u/34024373?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jonaskratochvil",
"html_url": "https://github.com/jonaskratochvil",
"followers_url": "https://api.github.com/users/jonaskratochvil/followers",
"following_url": "https://api.github.com/users/jonaskratochvil/following{/other_user}",
"gists_url": "https://api.github.com/users/jonaskratochvil/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jonaskratochvil/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jonaskratochvil/subscriptions",
"organizations_url": "https://api.github.com/users/jonaskratochvil/orgs",
"repos_url": "https://api.github.com/users/jonaskratochvil/repos",
"events_url": "https://api.github.com/users/jonaskratochvil/events{/privacy}",
"received_events_url": "https://api.github.com/users/jonaskratochvil/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"I would also be interested in this",
"I am interested in this as well. \r\nFor now, I'm using a simple approach by averaging the confidence score of each letter per word, resulting in a confidence score per word.\r\nYou can use [parlance](https://github.com/parlance/ctcdecode) to get a timestamp per letter with the beam search.",
"> I am interested in this as well.\r\n> For now, I'm using a simple approach by averaging the confidence score of each letter per word, resulting in a confidence score per word.\r\n> You can use [parlance](https://github.com/parlance/ctcdecode) to get a timestamp per letter with the beam search.\r\n\r\nHow do you get the confidence score of each letter?",
"> > I am interested in this as well.\r\n> > For now, I'm using a simple approach by averaging the confidence score of each letter per word, resulting in a confidence score per word.\r\n> > You can use [parlance](https://github.com/parlance/ctcdecode) to get a timestamp per letter with the beam search.\r\n> \r\n> How do you get the confidence score of each letter?\r\naveraging the score config score per letter\r\n",
"We don't support this yet, but we will let you know if we add support in the future. "
] | 2020-11-17T19:39:36 | 2021-08-17T00:56:58 | 2021-08-17T00:56:58 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello,
I am interested in getting the word level confidences from Quartznet model. What would be the best approach to obtain some proxy values to word level confidences? Taking the logit sequences and normalizing them by length seems like the most straight forward approach but I am not sure how reliable this would be. Any suggestions or ideas would be appreciated. | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1463/reactions",
"total_count": 5,
"+1": 5,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1463/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1462 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1462/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1462/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1462/events | https://github.com/NVIDIA/NeMo/issues/1462 | 744,770,813 | MDU6SXNzdWU3NDQ3NzA4MTM= | 1,462 | Possible problems with weight initialization in NeMo ASR[Question] | {
"login": "rbracco",
"id": 47190785,
"node_id": "MDQ6VXNlcjQ3MTkwNzg1",
"avatar_url": "https://avatars.githubusercontent.com/u/47190785?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rbracco",
"html_url": "https://github.com/rbracco",
"followers_url": "https://api.github.com/users/rbracco/followers",
"following_url": "https://api.github.com/users/rbracco/following{/other_user}",
"gists_url": "https://api.github.com/users/rbracco/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rbracco/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rbracco/subscriptions",
"organizations_url": "https://api.github.com/users/rbracco/orgs",
"repos_url": "https://api.github.com/users/rbracco/repos",
"events_url": "https://api.github.com/users/rbracco/events{/privacy}",
"received_events_url": "https://api.github.com/users/rbracco/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This is a good observation. I'll run some full scale experiments regarding this and respond in a day or two, if you don't mind. For clarification, when you say \"pytorch default\" - you mean you do not use the the weight initialization method at all and just use the default initialized weights that pytorch provides for Convs, yes?",
"Thanks for taking the time to look into it. Yes that's what I mean by pytorch default. It can be done either by commenting out `self.apply(lambda x: init_weights(x, mode=init_mode))` in the decoder, or overwriting it manually with `quartznet.decoder.decoder_layers[0] = nn.Conv1d(1024, <N_CLASSES>, kernel_size=1, stride=1)`. \r\n\r\nThe default init for a 1D conv in PyTorch is kaiming uniform, but I get a different standard deviation when using the PyTorch version and the NeMo version but I didn't dig too deep on why that might be. ",
"I think I might have an idea as to why applying the default is different as compared to applying `kaiming_uniform`\r\n\r\nThis is the `default` implementation of pytorch for all convND\r\n\r\n```python\r\n def reset_parameters(self) -> None:\r\n init.kaiming_uniform_(self.weight, a=math.sqrt(5))\r\n if self.bias is not None:\r\n fan_in, _ = init._calculate_fan_in_and_fan_out(self.weight)\r\n bound = 1 / math.sqrt(fan_in)\r\n init.uniform_(self.bias, -bound, bound)\r\n```\r\n\r\nNote the `a=sqrt(5)` as the param and the default `nonlinearity` value of `leaky_relu`.\r\n\r\nFor the `kaiming_uniform` mode in nemo - we compute the gain using the `relu` activation - as expected, if we dive deeper into what this gain value is actually computed into it can be found as\r\n```python\r\nnn.init.kaiming_uniform_(m.weight, nonlinearity=\"relu\") \r\n```\r\n\r\nwhich resolves to `a=0` and different `nonlinearity`\r\n```python\r\ndef kaiming_uniform_(tensor, a=0, mode='fan_in', nonlinearity='leaky_relu'):\r\n ...\r\n fan = _calculate_correct_fan(tensor, mode)\r\n gain = calculate_gain(nonlinearity, a)\r\n std = gain / math.sqrt(fan)\r\n bound = math.sqrt(3.0) * std # Calculate uniform bounds from standard deviation\r\n with torch.no_grad():\r\n return tensor.uniform_(-bound, bound)\r\n```\r\n\r\nHerein lies the difference in the gain computation\r\n```python\r\ndef calculate_gain(nonlinearity, param=None):\r\n ...\r\n linear_fns = ['linear', 'conv1d', 'conv2d', 'conv3d', 'conv_transpose1d', 'conv_transpose2d', 'conv_transpose3d']\r\n if nonlinearity in linear_fns or nonlinearity == 'sigmoid':\r\n return 1\r\n elif nonlinearity == 'tanh':\r\n return 5.0 / 3\r\n elif nonlinearity == 'relu':\r\n return math.sqrt(2.0)\r\n elif nonlinearity == 'leaky_relu':\r\n if param is None:\r\n negative_slope = 0.01\r\n elif not isinstance(param, bool) and isinstance(param, int) or isinstance(param, float):\r\n # True/False are instances of int, hence check above\r\n negative_slope = param\r\n else:\r\n raise ValueError(\"negative_slope {} not a valid number\".format(param))\r\n return math.sqrt(2.0 / (1 + negative_slope ** 2))\r\n else:\r\n raise ValueError(\"Unsupported nonlinearity {}\".format(nonlinearity))\r\n```\r\n\r\nNow, lets manually compute the output of the `compute_gain` method for `default` and `kaiming_uniform` init_mode in nemo\r\n\r\n`default` = compute_gain('leaky_relu', param=sqrt(5)) = sqrt(2.0 / (1. + 5.)) = sqrt(1./3.)\r\n`kaiming_uniform` = compute_gain('relu', param=0) = sqrt(2.0)\r\n\r\nThis is the reason the value of `default` does not match `kaiming_uniform`.",
"Very interesting! Thanks for looking into it. I am going to dive deeper and try to figure out why my model is performing better using pytorch defaults (it could still be chance) and I'll make sure to report back. \r\n\r\nAlso, if the decoder is a single layer and doesn't have an activation function (I guess softmax is the activation function), and our encoder is frozen, why do we need to init using kaiming? Shouldn't we just init to have a mean of 0 and unitary std dev? ",
"Hey @rbracco, I have some preliminary results (at least for from-scratch training). \r\n\r\n\r\n\r\n\r\nThough i havent plotted it here, the WERs (train, dev, test) exactly match the shape of the graph here. While this is just a point sample observation, xavier for both enc-dev is the best bet for from scratch training.\r\n\r\nThis doesnt invalidate your observation that loss reduces faster for finetuning. For the time being, I think we can enable a None flag for `init_mode` which would enable default pytorch initialization since it is a potential use case. We'll need extensive experimentation to show that its worth doing that for the decoder during finetuning however (which I currently can't do).\r\n",
"Related PR https://github.com/NVIDIA/NeMo/pull/1472",
"Awesome work @titu1994! Good to see that it isn't a problem for training from scratch. I'm not sure but maybe batchnorm lessens the importance of init since they all seem to end up in the same place. I will keep experimenting with transfer learning and report back. I just switched my training from English to Spanish with a totally different dataset and vocab, so I will try several inits on the new set and see if it is similar to what I experienced before, or if it was just a fluke. I should be able to report back early next week. ",
"PR https://github.com/NVIDIA/NeMo/pull/1472 is merged, and therefore you can now simply pass init_mode=None to get default pytorch initialization. ",
"Thank you! Just wanted to report back that in Spanish I failed to converge and then switched the init to be pytorch default and converged. I haven't had time to run full experiments, but I will probably do a writeup on transfer learning with NeMo at some point as I'm discovering lots of stuff that is causing faster convergence. "
] | 2020-11-17T14:10:36 | 2021-06-19T04:58:44 | 2021-06-19T04:58:44 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | ## Describe your question
1. Weight initialization in the `ConvASREncoder` and `ConvASRDecoder` defaults to `xavier_uniform` but the architectures use ReLU which does best with kaiming initialization. Why was xavier initialization chosen?
2. `ConvASREncoder` and `ConvASRDecoder` have an `init_mode` argument that delegates to `nemo.collections.asr.parts.jasper.init_weights` which returns different results than PyTorch's nn.init (the weights have a different initial standard deviation) and results in significantly worse training during transfer learning in my experiments. Why aren't PyTorch defaults used?
## Experimental Results**
I tried transfer learning from quartznet to a dataset with a different vocab, experimenting with 1 or 2 linear layers in my ASR decoder (decoder layer code included at end of post). I tried initializing decoder weights using NeMo's `xavier_uniform`, NeMo's `kaiming_uniform` and pytorch defaults (kaiming uniform is the default for 1d convs). I ran 12 trials for each. 6 were 2 epochs, and 6 were 1 epoch. LR=1e-3 (0.001). Mean loss after 1ep, and 2ep is included below.
### PyTorch Kaiming Uniform, mean loss 1ep=289.9, 2ep=179.9
### NeMo Kaiming Uniform: mean loss 1ep=515.5, 2ep=581.6
### NeMo Xavier Uniform (default): mean loss 1ep=566.8, 2ep=408.0
### Standard Deviation of weights after initialization
**2 Layer Decoder**:
**1 Layer Decoder**:
## Environment Details
Colab Pip Install - `pip install nemo-toolkit[all]==1.0.0b1`
Python 3.6.9
Pytorch 1.7
OS: Ubuntu 18.04.5 LTS
## Additional Details
### Definition of decoders
**2 Layer Decoder**
```
N_HIDDEN = 256
self.decoder_layers = torch.nn.Sequential(
torch.nn.Conv1d(self._feat_in, N_HIDDEN, kernel_size=1, bias=True),
torch.nn.ReLU(),
torch.nn.Conv1d(N_HIDDEN, self._num_classes, kernel_size=1, bias=True),
)
```
**1 Layer Decoder**
```
self.decoder_layers = torch.nn.Sequential(
torch.nn.Conv1d(self._feat_in, self._num_classes, kernel_size=1, bias=True),
)
``` | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1462/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1462/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1461 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1461/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1461/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1461/events | https://github.com/NVIDIA/NeMo/pull/1461 | 744,324,545 | MDExOlB1bGxSZXF1ZXN0NTIyMDcxNzM1 | 1,461 | Patch mp Jenkins test | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-17T01:10:47 | 2020-12-28T19:51:34 | 2020-11-17T01:14:48 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1461",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1461",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1461.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1461.patch",
"merged_at": "2020-11-17T01:14:48"
} | Line needed in Jenkins test lost in merge from v1.0.0b2. | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1461/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1461/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1460 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1460/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1460/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1460/events | https://github.com/NVIDIA/NeMo/pull/1460 | 744,280,985 | MDExOlB1bGxSZXF1ZXN0NTIyMDM2MDU1 | 1,460 | Bug fixes for text classification model | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-16T23:25:22 | 2020-11-18T20:15:24 | 2020-11-18T20:15:23 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1460",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1460",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1460.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1460.patch",
"merged_at": "2020-11-18T20:15:23"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1460/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1460/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1459 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1459/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1459/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1459/events | https://github.com/NVIDIA/NeMo/issues/1459 | 744,274,661 | MDU6SXNzdWU3NDQyNzQ2NjE= | 1,459 | [Question] Jasper/Quartznet ASR - Recommended length per file is between 5 and 25 seconds. Why? | {
"login": "ShantanuNair",
"id": 8559347,
"node_id": "MDQ6VXNlcjg1NTkzNDc=",
"avatar_url": "https://avatars.githubusercontent.com/u/8559347?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ShantanuNair",
"html_url": "https://github.com/ShantanuNair",
"followers_url": "https://api.github.com/users/ShantanuNair/followers",
"following_url": "https://api.github.com/users/ShantanuNair/following{/other_user}",
"gists_url": "https://api.github.com/users/ShantanuNair/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ShantanuNair/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ShantanuNair/subscriptions",
"organizations_url": "https://api.github.com/users/ShantanuNair/orgs",
"repos_url": "https://api.github.com/users/ShantanuNair/repos",
"events_url": "https://api.github.com/users/ShantanuNair/events{/privacy}",
"received_events_url": "https://api.github.com/users/ShantanuNair/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Not an expert, but I'll try to help (assuming you're talking about ASR). I believe the recommendation is due to most of the training data being in that range. Since the ASR models, Quartznet and Jasper, are fully convolutional, there's no RNN and I believe this means longer than 25s shouldn't cause too much of a problem, but give it a shot and see what kind of word error rate/accuracy you get.\r\n\r\nThe potential problems I see with...\r\n**Clips < 5 seconds** \r\nNot having enough context may cause problems. The receptive field of the convolutional ASR models is at least a few seconds, so really short bursts may cause a lot of the input to the model to be zeros (padding) resulting in weird results. Not sure if this actually happens\r\n\r\n**Clips > 25s**\r\nI would guess this isn't a problem for cpu inference. On GPU (inference, and especially training) you'd be limited to very small batch sizes and that could potentially cause problems. \r\n\r\nHope that helps and holds you over until a more experienced contributors is able to answer your question. ",
"@rbracco Thank you!! That's what I thought regarding a lack of RNN meaning >25s shouldn't be a problem. I ran with it and it seems okay, but still giving rather unusable output. I haven't run any issues with batch size since I'm testing out transcriptions with a single audio file using the pretrained QuartzNet15x5Base-En and haven't done any additional training yet. \r\n\r\n```\r\noh i meat five workers yesrano it's crazy so that's what i' saying it depends on what you consider my team um my research team within nhat product there's two of us that are like um full on workers and then we have some lab assistants that help out ofon the brigger product\r\n```\r\n\r\nSeems like beam-searching via a Language Model would help but I feel the quality is still lacking even without an LM. \r\n\r\n@okuchaiev Any comments regarding the original question? ",
"A bit late, but better late than never. The above comment is correct, Clips over 20 seconds of length would exhaust GPU memory, forcing tiny batch sizes and reducing training efficiency. Clips of any duration less than that is fine, but if its too short, then either the model wont have enough context to emit tokens."
] | 2020-11-16T23:10:45 | 2021-06-19T04:58:05 | 2021-06-19T04:58:05 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe your question**
Recommended length per file is between 5 and 25 seconds for most models. I'm curious where/how we arrive at this limitation, and what this means for results I get from running model.transcribe() on a audio longer than 25s. Some insight into where/why we have this limitation would be appreciated.
**Environment overview (please complete the following information)**
- Environment location: [Cloud(Collab)]
- Method of NeMo install: [from git with branch == v1.0.0b2]
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1459/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1459/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1458 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1458/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1458/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1458/events | https://github.com/NVIDIA/NeMo/issues/1458 | 743,941,672 | MDU6SXNzdWU3NDM5NDE2NzI= | 1,458 | RNN-T decoding: question about greedy batch inference | {
"login": "Oktai15",
"id": 17337773,
"node_id": "MDQ6VXNlcjE3MzM3Nzcz",
"avatar_url": "https://avatars.githubusercontent.com/u/17337773?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Oktai15",
"html_url": "https://github.com/Oktai15",
"followers_url": "https://api.github.com/users/Oktai15/followers",
"following_url": "https://api.github.com/users/Oktai15/following{/other_user}",
"gists_url": "https://api.github.com/users/Oktai15/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Oktai15/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Oktai15/subscriptions",
"organizations_url": "https://api.github.com/users/Oktai15/orgs",
"repos_url": "https://api.github.com/users/Oktai15/repos",
"events_url": "https://api.github.com/users/Oktai15/events{/privacy}",
"received_events_url": "https://api.github.com/users/Oktai15/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Yes, this is correct. I didn't think about reusing blank_mask for the label decoding step. Would you like to submit a PR to improve this for both greedy decoding implementations?"
] | 2020-11-16T15:44:18 | 2020-11-21T09:11:01 | 2020-11-21T09:11:01 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | As `blank_mask` contains ones where `time_idx >= out_len` and `k == self._blank_index` and zeros otherwise, we can replace the following code:
https://github.com/NVIDIA/NeMo/blob/5b6dd3850129898be47cf0d65587897ec45a5b59/nemo/collections/asr/parts/rnnt_greedy_decoding.py#L451-L454
with the following code (without masked_fill):
```python
for kidx, ki in enumerate(k):
if blank_mask[kidx] == 0:
label[kidx].append(ki)
```
It allows to remove `time_mask` at all. The following block:
https://github.com/NVIDIA/NeMo/blob/5b6dd3850129898be47cf0d65587897ec45a5b59/nemo/collections/asr/parts/rnnt_greedy_decoding.py#L391-L392
Can be squeezed to:
```python
blank_mask = time_idx >= out_len
```
Is it correct? @titu1994
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1458/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1458/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1457 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1457/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1457/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1457/events | https://github.com/NVIDIA/NeMo/issues/1457 | 743,711,829 | MDU6SXNzdWU3NDM3MTE4Mjk= | 1,457 | [Question] How to solve Exception while using another wav file: RuntimeError: Argument #4: Padding size should be less than the corresponding input dimension, but got: padding (256, 256) at dimension 2 of input [1, 1, 2] ? | {
"login": "sankulka",
"id": 22370827,
"node_id": "MDQ6VXNlcjIyMzcwODI3",
"avatar_url": "https://avatars.githubusercontent.com/u/22370827?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sankulka",
"html_url": "https://github.com/sankulka",
"followers_url": "https://api.github.com/users/sankulka/followers",
"following_url": "https://api.github.com/users/sankulka/following{/other_user}",
"gists_url": "https://api.github.com/users/sankulka/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sankulka/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sankulka/subscriptions",
"organizations_url": "https://api.github.com/users/sankulka/orgs",
"repos_url": "https://api.github.com/users/sankulka/repos",
"events_url": "https://api.github.com/users/sankulka/events{/privacy}",
"received_events_url": "https://api.github.com/users/sankulka/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@sankulka few questions:\r\n\r\n1) Are you able to successfully execute this notebook https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0b2/tutorials/NeMo_voice_swap_app.ipynb \r\n2) If you replace your file with this one https://dldata-public.s3.us-east-2.amazonaws.com/2086-149220-0033.wav , does the error go away?\r\n3) Do you know how many channels and what sample rate is your file? (Should be single channel 16Khz)",
"I had the same error. It was due to my microphone being stereo (2 channel) and 44.1Khz instead of mono (1 channel) and 16Khz as required. \r\n\r\nYou can check the sample_rate and resample if needed using torchaudio\r\n\r\n```\r\nimport torchaudio\r\n\r\ny, sr = torchaudio.load('my_sample.wav')\r\ny = y.mean(dim=0) # if there are multiple channels, average them to single channel\r\nif sr != 16000:\r\n resampler = torchaudio.transforms.Resample(sr, 16000)\r\n y_resampled = resampler(y)\r\ntorchaudio.save('my_sample_resampled.wav', y, sr)\r\n\r\nfiles = ['my_sample_resampled.wav']\r\nfor fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):\r\nprint(f\"Audio in {fname} was recognized as: {transcription}\")\r\n```",
"Thanks rbracco. Yes, by changing the sample rate, it worked well. Regards.",
"Yes, by changing the sample rate, it processed well. Thanks Robert.\n\nRegards,\nSantosh\n\nOn Thu, Nov 19, 2020 at 12:09 AM Robert Bracco <[email protected]>\nwrote:\n\n> I had the same error. It was due to my microphone being stereo (2 channel)\n> and 44.1Khz instead of mono (1 channel) and 16Khz as required.\n>\n> You can check the sample_rate and resample if needed using torchaudio\n>\n> import torchaudio\n>\n> y, sr = torchaudio.load('my_sample.wav')\n> y = y.mean(dim=0) # if there are multiple channels, average them to single channel\n> if sr != 16000:\n> resampler = torchaudio.transforms.Resample(sr, 16000)\n> y_resampled = resampler(y)\n> torchaudio.save('my_sample_resampled.wav')\n>\n> files = ['my_sample_resampled.wav']\n> for fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):\n> print(f\"Audio in {fname} was recognized as: {transcription}\")\n>\n> —\n> You are receiving this because you were mentioned.\n> Reply to this email directly, view it on GitHub\n> <https://github.com/NVIDIA/NeMo/issues/1457#issuecomment-729876934>, or\n> unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/AFKVUCYXGMDJHEGTWK6P4KDSQQIGBANCNFSM4TW7FMZQ>\n> .\n>\n",
"For me it's not working. showing an error:\r\n\r\n**save() missing 2 required positional arguments: 'src' and 'sample_rate'**\r\n\r\nPlease help me solve this problem",
"Can you post code? [torchaudio.save()](https://pytorch.org/audio/stable/torchaudio.html#i-o-functionalities) requires 3 arguments, a filepath, the audio, and the audio's sample_rate. I've edited the code above to include all 3.\r\n\r\nIt seems like you are doing something like `torchaudio.save('my_sample_resampled.wav')` but that's just the filepath, the correct would be `torchaudio.save('my_sample_resampled.wav', y, sr)`",
"Hi rbacco,\r\nThanks for your reply. I have solved this problem in different way. Instead of changing my recorded file I have changed my code for recording audio. Earlier I was using 2 channel while recording. When I changed it to 1 then all thing working fine. Code to record audio which have only one channel give below which is copied from https://dsp.stackexchange.com/questions/13728/what-are-chunks-when-recording-a-voice-signal\r\n\r\n**One channel recording:** \r\n\r\nimport pyaudio\r\nimport wave\r\nimport sys\r\n\r\nCHUNK = 1024\r\nWhat is CHUNKS here ?\r\nFORMAT = pyaudio.paInt16\r\nCHANNELS = 1\r\nRATE = 44100\r\nRECORD_SECONDS = 5\r\nWAVE_OUTPUT_FILENAME = \"my_sample.wav\"\r\n\r\np = pyaudio.PyAudio()\r\n\r\nstream = p.open(format=FORMAT,\r\n channels=CHANNELS,\r\n rate=RATE,\r\n input=True,\r\n frames_per_buffer=CHUNK)\r\nprint(\"start....\")\r\n\r\nframes = []\r\n\r\nfor i in range(0, int(RATE / CHUNK * RECORD_SECONDS)):\r\n data = stream.read(CHUNK)\r\n frames.append(data)\r\n\r\nprint(\"done...\")\r\n\r\n\r\nstream.stop_stream()\r\nstream.close()\r\np.terminate()\r\n\r\nwf = wave.open(WAVE_OUTPUT_FILENAME, 'wb')\r\nwf.setnchannels(CHANNELS)\r\nwf.setsampwidth(p.get_sample_size(FORMAT))\r\nwf.setframerate(RATE)\r\nwf.writeframes(b''.join(frames))\r\nwf.close()\r\n\r\nWhile prediction you can simply use:\r\nfiles = ['my_sample.wav']\r\nfor fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):\r\n print(f\"Audio in {fname} was recognized as: {transcription}\")\r\n\r\n**If you have audio file with two channel:**\r\n\r\nI have also found a different way to solve this problem. Error is created due to having two channel while recording , So we can take just one channel because both channel have minute change. I have change my code which provide me accurate result. Code is given below.\r\n\r\nimport torchaudio\r\nimport torch\r\n\r\ny, sr = torchaudio.load('my_sample.wav')\r\ny = torch.reshape(y[0], (1, y[0].size(0)))\r\ntorchaudio.save('my_sample_resampled.wav',y ,sr)\r\n\r\nfiles = ['my_sample_resampled.wav']\r\nfor fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):\r\n print(f\"Audio in {fname} was recognized as: {transcription}\")\r\n\r\n\r\n**Or you can take average both channel and convert it into one channel.**\r\n\r\nimport torchaudio\r\n\r\ny, sr = torchaudio.load('/content/output.wav')\r\ny = y.mean(dim = 0) # if there are multiple channels, average them to single channel\r\ny = torch.reshape(y,(1, y.size(0)))\r\ntorchaudio.save('my_sample_resampled.wav', y, sr)\r\n\r\nfiles = ['my_sample_resampled.wav']\r\nfor fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):\r\n print(f\"Audio in {fname} was recognized as: {transcription}\")",
"For me this error was being generated because the wav file had stereo channels. I needed to convert the file to mono channel:\r\n\r\n```\r\nfrom pydub import AudioSegment\r\nfile_path = \"input_sound_file.wav\"\r\nsound = AudioSegment.from_wav(file_path)\r\nsound = sound.set_channels(1)\r\nsound.export(file_path, format=\"wav\")\r\n\r\n```"
] | 2020-11-16T10:21:17 | 2022-01-02T10:30:42 | 2020-11-19T08:09:39 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe your question**
I just started off learning Nemo for ASR activities and getting exception if I send a different wav file to convert into text. Could you please share what pre-processing has to be performed for any other different wav file/format than an4 dataset
**A clear and concise description of your question.**
Describe what you want to achieve. And/or what NeMo APIs are unclear/confusing.
I am trying to send a wav file of < 20 sec duration to get the text output from the quartznet model. Here is a sample code:
files = ['my_sample.wav']
for fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):
print(f"Audio in {fname} was recognized as: {transcription}")
After this, I get below exception.
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
<ipython-input-53-f51e6e675965> in <module>()
1 files = ['my_sample.wav']
----> 2 for fname, transcription in zip(files, quartznet.transcribe(paths2audio_files=files)):
3 print(f"Audio in {fname} was recognized as: {transcription}")
14 frames
/usr/local/lib/python3.6/dist-packages/torch/autograd/grad_mode.py in decorate_context(*args, **kwargs)
24 def decorate_context(*args, **kwargs):
25 with self.__class__():
---> 26 return func(*args, **kwargs)
27 return cast(F, decorate_context)
28
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/models/ctc_models.py in transcribe(self, paths2audio_files, batch_size, logprobs)
158 for test_batch in temporary_datalayer:
159 logits, logits_len, greedy_predictions = self.forward(
--> 160 input_signal=test_batch[0].to(device), input_signal_length=test_batch[1].to(device)
161 )
162 if logprobs:
/usr/local/lib/python3.6/dist-packages/nemo/core/classes/common.py in __call__(self, wrapped, instance, args, kwargs)
509
510 # Call the method - this can be forward, or any other callable method
--> 511 outputs = wrapped(*args, **kwargs)
512
513 instance._attach_and_validate_output_types(output_types=output_types, out_objects=outputs)
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/models/ctc_models.py in forward(self, input_signal, input_signal_length, processed_signal, processed_signal_length)
394 if not has_processed_signal:
395 processed_signal, processed_signal_length = self.preprocessor(
--> 396 input_signal=input_signal, length=input_signal_length,
397 )
398
/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
/usr/local/lib/python3.6/dist-packages/nemo/core/classes/common.py in __call__(self, wrapped, instance, args, kwargs)
509
510 # Call the method - this can be forward, or any other callable method
--> 511 outputs = wrapped(*args, **kwargs)
512
513 instance._attach_and_validate_output_types(output_types=output_types, out_objects=outputs)
/usr/local/lib/python3.6/dist-packages/torch/autograd/grad_mode.py in decorate_context(*args, **kwargs)
24 def decorate_context(*args, **kwargs):
25 with self.__class__():
---> 26 return func(*args, **kwargs)
27 return cast(F, decorate_context)
28
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/modules/audio_preprocessing.py in forward(self, input_signal, length)
77 @torch.no_grad()
78 def forward(self, input_signal, length):
---> 79 processed_signal, processed_length = self.get_features(input_signal, length)
80
81 return processed_signal, processed_length
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/modules/audio_preprocessing.py in get_features(self, input_signal, length)
247
248 def get_features(self, input_signal, length):
--> 249 return self.featurizer(input_signal, length)
250
251 @property
/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py in _call_impl(self, *input, **kwargs)
725 result = self._slow_forward(*input, **kwargs)
726 else:
--> 727 result = self.forward(*input, **kwargs)
728 for hook in itertools.chain(
729 _global_forward_hooks.values(),
/usr/local/lib/python3.6/dist-packages/torch/autograd/grad_mode.py in decorate_context(*args, **kwargs)
24 def decorate_context(*args, **kwargs):
25 with self.__class__():
---> 26 return func(*args, **kwargs)
27 return cast(F, decorate_context)
28
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/parts/features.py in forward(self, x, seq_len)
345 # disable autocast to get full range of stft values
346 with torch.cuda.amp.autocast(enabled=False):
--> 347 x = self.stft(x)
348
349 # torch returns real, imag; so convert to magnitude
/usr/local/lib/python3.6/dist-packages/nemo/collections/asr/parts/features.py in <lambda>(x)
273 win_length=self.win_length,
274 center=True,
--> 275 window=self.window.to(dtype=torch.float),
276 )
277
/usr/local/lib/python3.6/dist-packages/torch/functional.py in stft(input, n_fft, hop_length, win_length, window, center, pad_mode, normalized, onesided, return_complex)
511 extended_shape = [1] * (3 - signal_dim) + list(input.size())
512 pad = int(n_fft // 2)
--> 513 input = F.pad(input.view(extended_shape), (pad, pad), pad_mode)
514 input = input.view(input.shape[-signal_dim:])
515 return _VF.stft(input, n_fft, hop_length, win_length, window, # type: ignore
/usr/local/lib/python3.6/dist-packages/torch/nn/functional.py in _pad(input, pad, mode, value)
3557 assert len(pad) == 2, '3D tensors expect 2 values for padding'
3558 if mode == 'reflect':
-> 3559 return torch._C._nn.reflection_pad1d(input, pad)
3560 elif mode == 'replicate':
3561 return torch._C._nn.replication_pad1d(input, pad)
RuntimeError: Argument #4: Padding size should be less than the corresponding input dimension, but got: padding (256, 256) at dimension 2 of input [1, 1, 2]
**Environment overview (please complete the following information)**
- Environment location: [Bare-metal, Docker, Cloud(specify cloud provider - AWS, Azure, GCP, Collab)]
Collab
- Method of NeMo install: [pip install or from source]. Please specify exact commands you used to install.
import nemo
import nemo.collections.asr as nemo_asr
- If method of install is [Docker], provide `docker pull` & `docker run` commands used
**Environment details**
If NVIDIA docker image is used you don't need to specify these.
Otherwise, please provide:
- OS version
- PyTorch version
- Python version
**Additional context**
Add any other context about the problem here.
Example: GPU model
| {
"login": "sankulka",
"id": 22370827,
"node_id": "MDQ6VXNlcjIyMzcwODI3",
"avatar_url": "https://avatars.githubusercontent.com/u/22370827?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sankulka",
"html_url": "https://github.com/sankulka",
"followers_url": "https://api.github.com/users/sankulka/followers",
"following_url": "https://api.github.com/users/sankulka/following{/other_user}",
"gists_url": "https://api.github.com/users/sankulka/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sankulka/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sankulka/subscriptions",
"organizations_url": "https://api.github.com/users/sankulka/orgs",
"repos_url": "https://api.github.com/users/sankulka/repos",
"events_url": "https://api.github.com/users/sankulka/events{/privacy}",
"received_events_url": "https://api.github.com/users/sankulka/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1457/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1457/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1456 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1456/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1456/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1456/events | https://github.com/NVIDIA/NeMo/issues/1456 | 743,023,559 | MDU6SXNzdWU3NDMwMjM1NTk= | 1,456 | [Question] I trained a model with Token_Classification_NER tutorial and saved it as .pt file how can I load up and use it in Pycharm? (or any python env to recommend) | {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"You may use the .pt file as any other Pytorch checkpoint or you can save your model as a .nemo file. The .nemo file will have everything needed (except datasets) to use your model. Please see the NeMo Primer tutorial for more information: https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0b2/tutorials/00_NeMo_Primer.ipynb#scrollTo=g91FE8mlMcnh",
"okay thank you!",
"I got the file as .nemo now, how can I just pass in a text file with only some sentences inside to have it predict the labels for me?\r\n\r\nAs it does in this part of the tutorial,\r\n`queries = [\r\n 'we bought four shirts from the nvidia gear store in santa clara.',\r\n 'Nvidia is a company.',\r\n 'The Adventures of Tom Sawyer by Mark Twain is an 1876 novel about a young boy growing '\r\n + 'up along the Mississippi River.',\r\n]\r\nresults = pretrained_ner_model.add_predictions(queries)\r\n\r\nfor query, result in zip(queries, results):\r\n print()\r\n print(f'Query : {query}')\r\n print(f'Result: {result.strip()}\\n')`",
"@181473h To load the pre-trained model:\r\nmodel = TokenClassificationModel.restore_from(\"path_to_.nemo\")\r\nmodel.add_predictions(queries)\r\n\r\nIf you need to continue training, please follow the setup in the example: https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/token_classification/token_classification.py#L89"
] | 2020-11-14T15:29:07 | 2020-11-23T05:08:53 | 2020-11-23T05:08:53 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | What are the imports or anything I may need? I have pip install nemo_toolkit[nlp] installed already.
Do I need to (2. Define and initialize the neural network) as mentioned in https://pytorch.org/tutorials/recipes/recipes/saving_and_loading_models_for_inference.html
How may I do it?
Appreciate any help T.T Thanks in advance
| {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1456/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1456/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1455 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1455/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1455/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1455/events | https://github.com/NVIDIA/NeMo/pull/1455 | 742,905,739 | MDExOlB1bGxSZXF1ZXN0NTIwOTQ2NDQ2 | 1,455 | Script for scoring ASR with sclite | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-14T02:29:51 | 2021-03-11T17:28:33 | 2020-11-18T22:32:54 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1455",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1455",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1455.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1455.patch",
"merged_at": "2020-11-18T22:32:54"
} | Also reinstated a script for processing hub5 data from NeMo 0.11 | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1455/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1455/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1454 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1454/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1454/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1454/events | https://github.com/NVIDIA/NeMo/pull/1454 | 742,865,367 | MDExOlB1bGxSZXF1ZXN0NTIwOTEyNDYy | 1,454 | speech commands tutorial with VAD support | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"> looks good to me. sorry for slow review\r\n\r\n@okuchaiev No worries! Thanks Oleksii!"
] | 2020-11-14T00:16:59 | 2021-01-22T15:29:24 | 2020-11-20T21:30:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1454",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1454",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1454.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1454.patch",
"merged_at": "2020-11-20T21:30:59"
} | I've tested on my local machine (macOS). Could you please help testing on your machine with microphone? @okuchaiev
Signed-off-by: fayejf <[email protected]> | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1454/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1454/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1453 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1453/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1453/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1453/events | https://github.com/NVIDIA/NeMo/pull/1453 | 742,604,122 | MDExOlB1bGxSZXF1ZXN0NTIwNjk2ODg5 | 1,453 | Added Pre-LN transformer encoder/decoder for NLP | {
"login": "MaximumEntropy",
"id": 9114321,
"node_id": "MDQ6VXNlcjkxMTQzMjE=",
"avatar_url": "https://avatars.githubusercontent.com/u/9114321?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MaximumEntropy",
"html_url": "https://github.com/MaximumEntropy",
"followers_url": "https://api.github.com/users/MaximumEntropy/followers",
"following_url": "https://api.github.com/users/MaximumEntropy/following{/other_user}",
"gists_url": "https://api.github.com/users/MaximumEntropy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MaximumEntropy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MaximumEntropy/subscriptions",
"organizations_url": "https://api.github.com/users/MaximumEntropy/orgs",
"repos_url": "https://api.github.com/users/MaximumEntropy/repos",
"events_url": "https://api.github.com/users/MaximumEntropy/events{/privacy}",
"received_events_url": "https://api.github.com/users/MaximumEntropy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "MaximumEntropy",
"id": 9114321,
"node_id": "MDQ6VXNlcjkxMTQzMjE=",
"avatar_url": "https://avatars.githubusercontent.com/u/9114321?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MaximumEntropy",
"html_url": "https://github.com/MaximumEntropy",
"followers_url": "https://api.github.com/users/MaximumEntropy/followers",
"following_url": "https://api.github.com/users/MaximumEntropy/following{/other_user}",
"gists_url": "https://api.github.com/users/MaximumEntropy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MaximumEntropy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MaximumEntropy/subscriptions",
"organizations_url": "https://api.github.com/users/MaximumEntropy/orgs",
"repos_url": "https://api.github.com/users/MaximumEntropy/repos",
"events_url": "https://api.github.com/users/MaximumEntropy/events{/privacy}",
"received_events_url": "https://api.github.com/users/MaximumEntropy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "MaximumEntropy",
"id": 9114321,
"node_id": "MDQ6VXNlcjkxMTQzMjE=",
"avatar_url": "https://avatars.githubusercontent.com/u/9114321?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MaximumEntropy",
"html_url": "https://github.com/MaximumEntropy",
"followers_url": "https://api.github.com/users/MaximumEntropy/followers",
"following_url": "https://api.github.com/users/MaximumEntropy/following{/other_user}",
"gists_url": "https://api.github.com/users/MaximumEntropy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MaximumEntropy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MaximumEntropy/subscriptions",
"organizations_url": "https://api.github.com/users/MaximumEntropy/orgs",
"repos_url": "https://api.github.com/users/MaximumEntropy/repos",
"events_url": "https://api.github.com/users/MaximumEntropy/events{/privacy}",
"received_events_url": "https://api.github.com/users/MaximumEntropy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [] | 2020-11-13T16:53:39 | 2021-01-22T15:29:14 | 2020-11-23T19:30:52 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1453",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1453",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1453.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1453.patch",
"merged_at": "2020-11-23T19:30:52"
} | Signed-off-by: MaximumEntropy <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1453/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1453/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1452 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1452/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1452/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1452/events | https://github.com/NVIDIA/NeMo/issues/1452 | 742,280,416 | MDU6SXNzdWU3NDIyODA0MTY= | 1,452 | [Question]Does anyone try to optimize quartznet inference through tensorrt? | {
"login": "zjyuan",
"id": 344545,
"node_id": "MDQ6VXNlcjM0NDU0NQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/344545?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zjyuan",
"html_url": "https://github.com/zjyuan",
"followers_url": "https://api.github.com/users/zjyuan/followers",
"following_url": "https://api.github.com/users/zjyuan/following{/other_user}",
"gists_url": "https://api.github.com/users/zjyuan/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zjyuan/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zjyuan/subscriptions",
"organizations_url": "https://api.github.com/users/zjyuan/orgs",
"repos_url": "https://api.github.com/users/zjyuan/repos",
"events_url": "https://api.github.com/users/zjyuan/events{/privacy}",
"received_events_url": "https://api.github.com/users/zjyuan/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-13T08:58:12 | 2021-07-13T15:20:12 | 2021-07-13T15:20:12 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe your question**
I follow [this link](https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0b2/tutorials/asr/01_ASR_with_NeMo.ipynb) to try to achieve quartznet inference optimization through tensorrt. I have installed onnxruntime-gpu-tensorrt and use tensorrt provider in onnxruntime. I did observe high gpu usage and gpu memory ocupation. But the inference time is a little slower than without tensorrt. I heard that some neural network can't be optimized by tensorrt. So is quartznet that case? Does anyone try to optimize quartznet inference through tensorrt? If yes, can you share some data with me? Thanks.
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1452/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1452/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1451 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1451/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1451/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1451/events | https://github.com/NVIDIA/NeMo/pull/1451 | 742,033,417 | MDExOlB1bGxSZXF1ZXN0NTIwMjQ1MzY5 | 1,451 | Add try/except for both ruamel.yaml and ruamel_yaml | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-13T00:43:14 | 2020-11-13T00:45:40 | 2020-11-13T00:45:34 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1451",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1451",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1451.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1451.patch",
"merged_at": "2020-11-13T00:45:34"
} | Signed-off-by: Jocelyn Huang <[email protected]> | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1451/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1451/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1450 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1450/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1450/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1450/events | https://github.com/NVIDIA/NeMo/pull/1450 | 742,027,299 | MDExOlB1bGxSZXF1ZXN0NTIwMjQwMzk3 | 1,450 | Dataset creation tool based on CTC-segmentation | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [] | 2020-11-13T00:27:07 | 2020-11-13T22:55:32 | 2020-11-13T22:55:23 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1450",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1450",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1450.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1450.patch",
"merged_at": "2020-11-13T22:55:23"
} | Adding a tool for an ASR dataset creation.
Long audio files and corresponding transcripts are segmented with the algorithm described in [CTC-Segmentation of Large Corpora for German End-to-end Speech Recognition](https://arxiv.org/pdf/2007.09127.pdf).
Provided functionality:
- a complete pipeline from audio and text pre-processing, including minimal normalization needed for proper alignment with [ctc_segmentation package](https://github.com/lumaku/ctc-segmentation) to manifest generation and verification utils
- support for parallel segmentation of multiple files
A tutorial on how to use the tool is included. | {
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1450/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1450/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1449 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1449/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1449/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1449/events | https://github.com/NVIDIA/NeMo/pull/1449 | 742,020,986 | MDExOlB1bGxSZXF1ZXN0NTIwMjM1MTMz | 1,449 | Switch ruamel.yaml to ruamel_yaml in notebook for compatibility with … | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-13T00:11:26 | 2020-11-13T00:34:46 | 2020-11-13T00:34:45 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1449",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1449",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1449.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1449.patch",
"merged_at": "2020-11-13T00:34:45"
} | …nvcr.io/nvidia/pytorch:20.09-py3 container
Signed-off-by: Jocelyn Huang <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1449/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1449/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1448 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1448/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1448/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1448/events | https://github.com/NVIDIA/NeMo/pull/1448 | 742,016,829 | MDExOlB1bGxSZXF1ZXN0NTIwMjMxODE5 | 1,448 | ONNX Runtime 1.5.2 | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-13T00:00:35 | 2020-11-13T00:04:08 | 2020-11-13T00:04:08 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1448",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1448",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1448.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1448.patch",
"merged_at": "2020-11-13T00:04:07"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1448/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1448/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1447 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1447/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1447/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1447/events | https://github.com/NVIDIA/NeMo/pull/1447 | 741,960,358 | MDExOlB1bGxSZXF1ZXN0NTIwMTg0NjM1 | 1,447 | Minor bug fix | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-12T22:04:20 | 2020-11-25T16:47:54 | 2020-11-13T00:35:17 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1447",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1447",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1447.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1447.patch",
"merged_at": "2020-11-13T00:35:17"
} | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1447/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1447/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1446 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1446/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1446/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1446/events | https://github.com/NVIDIA/NeMo/pull/1446 | 741,929,750 | MDExOlB1bGxSZXF1ZXN0NTIwMTU5MjYz | 1,446 | Minor bug fix | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-12T21:09:25 | 2020-11-25T16:50:13 | 2020-11-12T22:03:47 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1446",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1446",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1446.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1446.patch",
"merged_at": null
} | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1446/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1446/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1445 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1445/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1445/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1445/events | https://github.com/NVIDIA/NeMo/issues/1445 | 741,704,608 | MDU6SXNzdWU3NDE3MDQ2MDg= | 1,445 | How to write/save/load a simple custom decoder using nn.Sequential? [Question] | {
"login": "rbracco",
"id": 47190785,
"node_id": "MDQ6VXNlcjQ3MTkwNzg1",
"avatar_url": "https://avatars.githubusercontent.com/u/47190785?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rbracco",
"html_url": "https://github.com/rbracco",
"followers_url": "https://api.github.com/users/rbracco/followers",
"following_url": "https://api.github.com/users/rbracco/following{/other_user}",
"gists_url": "https://api.github.com/users/rbracco/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rbracco/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rbracco/subscriptions",
"organizations_url": "https://api.github.com/users/rbracco/orgs",
"repos_url": "https://api.github.com/users/rbracco/repos",
"events_url": "https://api.github.com/users/rbracco/events{/privacy}",
"received_events_url": "https://api.github.com/users/rbracco/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"You should not be overriding your decoder as such. You should use a custom neural module. \r\n\r\n1) Make your custom decoder extend neural module class\r\n2) Override the nemo.collections.asr.modules.ConvASRDecoder classpath with the classpath of your new decoder. \r\n3) Create the model with this config and train. \r\n\r\nDuring evaluation, make sure to import the custom decoder before you use restore from. ",
"Thank you, I will try this and follow up if I have any questions. If I get it working I will leave an example for future people with the same issue. ",
"Okay so I worked on this for a while and then got stuck. I wrote a TwoLayerDecoder class that extends NeuralModule but I'm not sure how to use it with a pretrained model. I have two questions.\r\n\r\n1. Once I load quartznet w pretrained weights, how do I then swap in my decoder using the config file, but keeping the pretrained weights in the encoder? \r\n2. What is the best way to achieve my two layer decoder? Should I write a new class extending NeuralModule and copy in most of the code from `nemo.collections.asr.modules.ConvASRDecoder` and then change what I need? Or is it okay to inherit from `nemo.collections.asr.modules.ConvASRDecoder` and just write a new `__init__` and `forward`? Thank you!",
"For changing the decoder without changing vocabulary size (keep 28 char vocab size of QN) it's done as follows\r\n\r\nFor 1) you should be able to do \r\n`quartznet.decoder = MyNewTwoLayerDecoder()`. \r\nDon't forget to update `quartzNet.cfg.decoder._target_` with the new class path to your file, otherwise restore from won't work. \r\n\r\nThat's about it. \r\n\r\nFor a decoder that changes the vocabulary size, first use the quartznet.change_vocabulary() method then do the above steps. \r\n\r\nFor 2) inheriting is fine, only if you override the decoder layer correctly so as to remove the previous created weights with your own. For clean approach, I'd say copy paste the code and edit the portions you wish to change. ",
"Thank you, so I actually tried 3 different ways...\r\n\r\n1. Copy code and edit then do `quartznet.decoder = MyNewTwoLayerDecoder(*args)`\r\n2. Inheritance\r\n3. Copy code and edit and then create a new model from config and do `quartznet.decoder=new_model.decoder` (this one mainly to make sure the config file would work)\r\n\r\nAll 3 methods successfully combine the encoder pretrained weights and the new decoder. I then freeze the encoder layers and fit, but in each case I get `KeyError: 30` for the line `reference = ''.join([self.labels_map[c] for c in target])` in `wer.py`. I have 41 classes (including blank label) and the correct vocab and # of classes appears when I run `quartznet.decoder.vocabulary` but when I enter the debugger the `labels_map` attribute of WER (a dict mapping int values to labels) is still using the original English alphabet. I did make sure to call `quartznet.change_vocabulary(new_vocabulary=my_new_vocab)` but somehow it doesn't fix it.\r\n\r\nAny ideas? Please let me know if this warrants a new issue and in the meantime I'll keep digging. Thanks so much for your time and help.",
"Oh wow, so 2 seconds after posting this, I reran the code but swapping the order of the execution to be \r\n\r\n```\r\nquartznet.change_vocabulary(new_vocabulary=my_new_vocab)\r\nquartznet.decoder = ConvASRDecoderTwo(1024, 40, vocabulary=my_new_vocab)\r\n```\r\n\r\ninstead of \r\n\r\n```\r\nquartznet.decoder = ConvASRDecoderTwo(1024, 40, vocabulary=my_new_vocab)\r\nquartznet.change_vocabulary(new_vocabulary=my_new_vocab)\r\n```\r\n\r\nand the KeyError disappeared. It appears that you need to change the vocabulary on the pretrained model prior to instantiating the new decoder. \r\n\r\nI'm still not sure everything is worked out because the loss is coming down much more slowly than when I overwrote the decoder manually with `quartznet.decoder.decoder_layers = nn.Sequential(nn.Conv1d(1024, 256, kernel_size=1, stride=1), nn.ReLU(), nn.Conv1d(256, 41, kernel_size=1, stride=1))`. I'll keep digging and report back. ",
"The error dissapears but actually now wer will be incorrect - it will assume the CTC blank I'd is 29, but you have 41 labels. \r\n\r\nI would rather suggest this (apologies for the roundabout way above) -\r\n\r\n1) create neural module\r\n2) change `QuartzNet.cfg.decoder._target_` to classpath of new decoder\r\n3) simply call change vocabulary. \r\n\r\nThat should be all that's actually needed. Please let me know If this works ",
"Thank you for this. I'm trying it now but am running into some issues setting the config. I started with code from the ASR with NeMo tutorial and they appear to use Hydra 1.0. I see that in Hydra 1.1 `cls` is replaced with `_target_` in the config files, so I followed your instructions but for step 2 I changed `quartznet.cfg.decoder.cls` to the classpath of my new decoder. When I call `change_vocabulary` however, my decoder doesn't change. \r\n\r\nExactly what I did is below, any ideas?\r\n\r\n### Step 1: Create Neural Module\r\nAll I changed here is A. Changed the decoder layers from a single 1d conv to 1dconv->relu->1dconv. B. Commented out the line that does weight initialization for Nemo.\r\n```\r\nclass ConvASRDecoderTwo(NeuralModule, Exportable):\r\n \"\"\"Simple ASR Decoder for use with CTC-based models such as JasperNet and QuartzNet\r\n Based on these papers:\r\n https://arxiv.org/pdf/1904.03288.pdf\r\n https://arxiv.org/pdf/1910.10261.pdf\r\n https://arxiv.org/pdf/2005.04290.pdf\r\n \"\"\"\r\n\r\n def save_to(self, save_path: str):\r\n pass\r\n\r\n @classmethod\r\n def restore_from(cls, restore_path: str):\r\n pass\r\n\r\n @property\r\n def input_types(self):\r\n return OrderedDict({\"encoder_output\": NeuralType(('B', 'D', 'T'), AcousticEncodedRepresentation())})\r\n\r\n @property\r\n def output_types(self):\r\n return OrderedDict({\"logprobs\": NeuralType(('B', 'T', 'D'), LogprobsType())})\r\n\r\n def __init__(self, feat_in, num_classes, init_mode=\"xavier_uniform\", vocabulary=None):\r\n super().__init__()\r\n if vocabulary is not None:\r\n if num_classes != len(vocabulary):\r\n raise ValueError(\r\n f\"If vocabulary is specified, it's length should be equal to the num_classes. Instead got: num_classes={num_classes} and len(vocabulary)={len(vocabulary)}\"\r\n )\r\n self.__vocabulary = vocabulary\r\n self._feat_in = feat_in\r\n # Add 1 for blank char\r\n self._num_classes = num_classes + 1\r\n\r\n self.decoder_layers = torch.nn.Sequential(\r\n torch.nn.Conv1d(self._feat_in, 256, kernel_size=1, bias=True),\r\n torch.nn.ReLU(),\r\n torch.nn.Conv1d(256, self._num_classes, kernel_size=1, bias=True),\r\n )\r\n #self.apply(lambda x: init_weights(x, mode=init_mode))\r\n\r\n @typecheck()\r\n def forward(self, encoder_output):\r\n return torch.nn.functional.log_softmax(self.decoder_layers(encoder_output).transpose(1, 2), dim=-1)\r\n\r\n def input_example(self):\r\n \"\"\"\r\n Generates input examples for tracing etc.\r\n Returns:\r\n A tuple of input examples.\r\n \"\"\"\r\n bs = 8\r\n seq = 64\r\n input_example = torch.randn(bs, self._feat_in, seq).to(next(self.parameters()).device)\r\n return tuple([input_example])\r\n\r\n def _prepare_for_export(self):\r\n m_count = 0\r\n for m in self.modules():\r\n if type(m).__name__ == \"MaskedConv1d\":\r\n m.use_mask = False\r\n m_count += 1\r\n if m_count > 0:\r\n logging.warning(f\"Turned off {m_count} masked convolutions\")\r\n Exportable._prepare_for_export(self)\r\n\r\n @property\r\n def vocabulary(self):\r\n return self.__vocabulary\r\n\r\n @property\r\n def num_classes_with_blank(self):\r\n return self._num_classes\r\n```\r\n\r\n## Step 2 and 3: change `QuartzNet.cfg.decoder.cls` to classpath of new decoder, call `change_vocabulary`\r\n\r\n\r\n"
] | 2020-11-12T15:40:11 | 2021-07-13T15:20:02 | 2021-07-13T15:20:02 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe your question**
I would like to implement my own decoder for transfer learning to experiment with multiple linear layers. I did so by overriding quartznet.decoder.decoder_layers with my own `nn.Sequential` using the code below
```
quartznet.decoder.decoder_layers = nn.Sequential(nn.Conv1d(1024, 256, kernel_size=1, stride=1), nn.ReLU(), nn.Conv1d(256, 41, kernel_size=1, stride=1))
```
This worked and it trains well, but when I save the model and try to load it, I get the following error:
```
RuntimeError: Error(s) in loading state_dict for EncDecCTCModel:
Unexpected key(s) in state_dict: "decoder.decoder_layers.2.weight", "decoder.decoder_layers.2.bias".
size mismatch for decoder.decoder_layers.0.weight: copying a param with shape torch.Size([256, 1024, 1]) from checkpoint, the shape in current model is torch.Size([41, 1024, 1]).
size mismatch for decoder.decoder_layers.0.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([41]).
```
This is because my underlying config file still specifies the decoder as being from the class `nemo.collections.asr.modules.ConvASRDecoder`. I have no idea how to update the config file to use my new decoder, or how to bypass the config file altogether, and couldn't find a way to do so in the docs. Even if I load up quartznet and manually overwrite the decoder, then try to load the saved checkpoint, it fails because it seems to be using the config file behind the scenes.
**Environment overview (please complete the following information)**
Colab using nemo-toolkit[all]==1.0.0b1 and config from https://raw.githubusercontent.com/NVIDIA/NeMo/main/examples/asr/conf/config.yaml
**Environment details**
If NVIDIA docker image is used you don't need to specify these.
Python 3.6.9
Pytorch 1.7
OS: Ubuntu 18.04.5 LTS
**Additional context**
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1445/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1445/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1444 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1444/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1444/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1444/events | https://github.com/NVIDIA/NeMo/issues/1444 | 741,692,258 | MDU6SXNzdWU3NDE2OTIyNTg= | 1,444 | How to get word level timestamps with External Language Model? | {
"login": "anujsuchchal",
"id": 46239398,
"node_id": "MDQ6VXNlcjQ2MjM5Mzk4",
"avatar_url": "https://avatars.githubusercontent.com/u/46239398?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anujsuchchal",
"html_url": "https://github.com/anujsuchchal",
"followers_url": "https://api.github.com/users/anujsuchchal/followers",
"following_url": "https://api.github.com/users/anujsuchchal/following{/other_user}",
"gists_url": "https://api.github.com/users/anujsuchchal/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anujsuchchal/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anujsuchchal/subscriptions",
"organizations_url": "https://api.github.com/users/anujsuchchal/orgs",
"repos_url": "https://api.github.com/users/anujsuchchal/repos",
"events_url": "https://api.github.com/users/anujsuchchal/events{/privacy}",
"received_events_url": "https://api.github.com/users/anujsuchchal/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@anujsuchchal Did you come across a solution for this?",
"@ShantanuNair Yes. For this I am using another CTC Beam Search Decoder package by [Parlance](https://github.com/parlance/ctcdecode). \r\nThis returns timesteps of characters as well in addition to Beam results which you can use to align Audio with Transcript. Hope this helps.\r\nClosing this issue now.\r\n\r\n\r\n",
"@ShantanuNair and @anujsuchchal I've tried CTC decoder package by parlance to get the time stamp for the output for conformer-CTC model. I'm getting output but unable to understand how to get the timestamp for each word. Can you guys help me with this?\r\n",
"@dbkalaria did you ever manage to figure out how to use the timestamp information to calculate word-level timestamps? I am stuck on the same problem too.",
"@abarcovschi Yes I was able to extract word-level timestamp information although it was not that accurate and I had tried that about 2 years ago, didn't know whether they had added that feature or not as there was no direct implementation at that time. \r\n\r\nFYI, I tried different options like vosk, whisper, etc and ended up using whisper model for my task.",
"I was able to extract word-level timestamp information in the following way:\r\n\r\nUsing `transcribe_speech.py` with the `compute_timestamps=True` argument, the `transcriptions` object is a list of `Hypothesis` objects that have a `timestep` instance variable, which is a dictionary that has a `word` field, which is a list of dictionary objects that contain the start and end timestep (offset) for each word in the transcription text.\r\n\r\nThe start and end offset for each word can be converted to the start and end time, in seconds, for each word in the audio file using the following code:\r\n\r\n```\r\n# create config object from config yaml file of the model.\r\ncfg = Config(os.path.join(args.model_folder_path, \"hparams.yaml\"))\r\n\r\n# get the seconds duration of a timestep at output of the Conformer.\r\ntime_stride = 4 * cfg.get(\"cfg/preprocessor/window_stride\")\r\n\r\n# for each word dict:\r\nstart_time = word['start_offset'] * time_stride\r\nstop_time = word['end_offset'] * time_stride\r\n```\r\n\r\n**NOTE:** I am using a Conformer model here, so the time_stride has a multiplication factor of 4, I am not sure if it is different for other types of ASR models.\r\n"
] | 2020-11-12T15:25:15 | 2023-12-16T17:05:20 | 2020-12-12T17:13:30 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | How can we get word level timestamps for the output words of external LM?
Since the word timestamps logic given in Offline_ASR.ipynb works on Greedy CTC decoder. Number of words in Greedy decoder and Beam Search decoder are usually different so we can't use this logic with Beam Search Decoder.
Is there any way to do this? I have seen some code related to timesteps in source code of ctcdecode package, do I need to try to implement greedy decoder logic into source code of ctcdecode to return timestamps and build it again? Or there is any function/package already available?
Please help. | {
"login": "anujsuchchal",
"id": 46239398,
"node_id": "MDQ6VXNlcjQ2MjM5Mzk4",
"avatar_url": "https://avatars.githubusercontent.com/u/46239398?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anujsuchchal",
"html_url": "https://github.com/anujsuchchal",
"followers_url": "https://api.github.com/users/anujsuchchal/followers",
"following_url": "https://api.github.com/users/anujsuchchal/following{/other_user}",
"gists_url": "https://api.github.com/users/anujsuchchal/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anujsuchchal/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anujsuchchal/subscriptions",
"organizations_url": "https://api.github.com/users/anujsuchchal/orgs",
"repos_url": "https://api.github.com/users/anujsuchchal/repos",
"events_url": "https://api.github.com/users/anujsuchchal/events{/privacy}",
"received_events_url": "https://api.github.com/users/anujsuchchal/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1444/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1444/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1443 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1443/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1443/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1443/events | https://github.com/NVIDIA/NeMo/issues/1443 | 741,507,561 | MDU6SXNzdWU3NDE1MDc1NjE= | 1,443 | [Question] Why 256 Epoch for German and 512 for Russian in Transfer learning (Based on Paper)? | {
"login": "soheiltehranipour",
"id": 37288139,
"node_id": "MDQ6VXNlcjM3Mjg4MTM5",
"avatar_url": "https://avatars.githubusercontent.com/u/37288139?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/soheiltehranipour",
"html_url": "https://github.com/soheiltehranipour",
"followers_url": "https://api.github.com/users/soheiltehranipour/followers",
"following_url": "https://api.github.com/users/soheiltehranipour/following{/other_user}",
"gists_url": "https://api.github.com/users/soheiltehranipour/gists{/gist_id}",
"starred_url": "https://api.github.com/users/soheiltehranipour/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/soheiltehranipour/subscriptions",
"organizations_url": "https://api.github.com/users/soheiltehranipour/orgs",
"repos_url": "https://api.github.com/users/soheiltehranipour/repos",
"events_url": "https://api.github.com/users/soheiltehranipour/events{/privacy}",
"received_events_url": "https://api.github.com/users/soheiltehranipour/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Can you share more info about your training process?\r\n\r\n-How many hours of training data do you have?\r\n-What is the size of the Persian vocab? Did you change the decoder to use the new vocab? \r\n-What is the loss doing? Is it falling to a certain level and stalling? What's the lowest loss you get and can you share a graph if possible? \r\n\r\nIf you have a colab notebook I'd be happy to take a look. I also discovered a possible flaw in the initialization of the decoder weights in NeMo as discussed in #1462. Using the standard initialization my transfer learning frequently fails to converge. \r\n\r\nI had more success by doing the following:\r\n\r\n```\r\nquartznet = nemo_asr.models.EncDecCTCModel.from_pretrained(model_name=\"QuartzNet15x5Base-En\")\r\nquartznet.change_vocabulary(new_vocabulary=en_ipa_vocab)\r\nprint(quartznet.decoder)\r\n```\r\n\r\n\r\n\r\n\r\nThen manually reinitialize the first layer of the decoder so that it uses PyTorch's default Kaiming Uniform initialization instead of NeMo's Xavier Uniform. Replace <N_CLASSES> with the number of symbols in your vocab (+1 for the blank token). It is the same value that has a red box around it in my decoder above. \r\n\r\n```\r\nquartznet.decoder.decoder_layers[0] = nn.Conv1d(1024, <N_CLASSES>, kernel_size=1, stride=1)\r\n```\r\n\r\nHope this helps and let me know if you have any questions. "
] | 2020-11-12T11:09:53 | 2021-05-28T08:45:35 | 2021-05-28T08:45:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Question
First of all, thank you guys specially @okuchaiev for this magnificent project on Conversational A.I. Good job Guys.
I am using NeMo **ASR** for **Persian** language mainly because of being able to use **transfer learning** due to lack of labeled data. I got 40% WER on my validation data after working on NeMo for about 3 months. For more information, I am using Quartznet15x5 (En) and used transfer learning for another language. _But there is 2 main question here:_
1. According to the [Paper ](https://arxiv.org/abs/2005.04290): "The Russian model was trained for 512 epochs while the Spanish and German ones for 256 epochs."
But I am stuck on the 40% WER after almost 15th epoch and it is not changing even after 50 epochs. **Am i experiencing overfitting here?**
On the other hand, based on the main use of transfer learning, there is no need to train the per-trained model for this **such large number of epochs.** Please let me know what is the reason of this strategy of training.
2. **If this is overfitting, How can i regularize it?**
(Can Dropout help me with that? ) If YES, is there any advice for setting that?
Environment overview:
-Environment location: Jupyter Notebook (Cloud Services)
-Method of NeMo install: !python -m pip install git+https://github.com/NVIDIA/NeMo.git@main#egg=nemo_toolkit[all]
Environment details
-OS version : Ubuntu 18.04
-PyTorch version : 1.6.0+cu101
-Python version : Python 3.6.9
-GPU model: GTX 1080 Ti | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1443/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1443/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1442 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1442/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1442/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1442/events | https://github.com/NVIDIA/NeMo/pull/1442 | 741,386,787 | MDExOlB1bGxSZXF1ZXN0NTE5NzA5NjIx | 1,442 | Added new features to Speech Data Explorer | {
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging 918f97f81514c0f243321e81eb73adcd49c18cc3 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-a817d9e013ff7a00f492166aeb5bf1830ad2cba6)\n\n**new alerts:**\n\n* 2 for Except block handles 'BaseException'"
] | 2020-11-12T08:22:30 | 2020-11-13T21:53:11 | 2020-11-13T21:53:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1442",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1442",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1442.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1442.patch",
"merged_at": "2020-11-13T21:53:07"
} | Added new features to Speech Data Explorer:
- backend paging/sorting/filtering (allows to work with large datasets)
- support for assigning a port
- minor tweaks to UI
Signed-off-by: Vitaly Lavrukhin <[email protected]> | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1442/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1442/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1441 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1441/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1441/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1441/events | https://github.com/NVIDIA/NeMo/pull/1441 | 741,279,776 | MDExOlB1bGxSZXF1ZXN0NTE5NjIwNDU0 | 1,441 | pip uninstall onnxruntime | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-12T04:58:30 | 2020-11-12T05:26:24 | 2020-11-12T05:26:24 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1441",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1441",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1441.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1441.patch",
"merged_at": "2020-11-12T05:26:24"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1441/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1441/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1440 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1440/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1440/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1440/events | https://github.com/NVIDIA/NeMo/pull/1440 | 741,099,027 | MDExOlB1bGxSZXF1ZXN0NTE5NDc0MTIw | 1,440 | Register BERT model config for Jarvis | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"awesome!!",
"This pull request **introduces 2 alerts** when merging 12122b266f59f46f76414748c73da675f830ec03 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-cf2265463b3732f561da98b7d2b7b43d39332d52)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging 46b256d87db223294478f98dac0619c4d964452a into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-9b15b496347258ec100ac86d6420295375459224)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging 176b240529c2007c0807ae1c46a053532b1dccd8 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-18f8083ea32be718223835aa8f576a38940e5d19)\n\n**new alerts:**\n\n* 2 for Unused import",
"Once everyone is happy with this PR I can make changes to other NLP models that need to register artifacts for Jarvis.",
"> @ericharper please change `nn_config.json` to ``bert_config.json` and we're good to go :)\r\n\r\nI vote for `pretrain_config.json`. \r\nRest LGTM.\r\n\r\n",
"Had a discussion with @tkornuta-nvidia and best if we go with \r\n`megatron_config.json` and `bert_config.json` for now.",
"This pull request **introduces 8 alerts** when merging 0b9e8446a6b69fc1f90d184897c5bb04d09acccc into fb802c524d784da616e7fe7788255ef89520682d - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-59ed07db28c65458b07356418e534e132426a4e9)\n\n**new alerts:**\n\n* 8 for Unused import",
"This pull request **introduces 8 alerts** when merging 6a376ba3ccd63a4ddc83f0cb9204f789e8bbd7ac into afdfe6879589d2c21aa73f84def1f53d9b8a3211 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-e5650899bbd05ec43832c812ad13807f1d87a2f9)\n\n**new alerts:**\n\n* 8 for Unused import",
"This pull request **introduces 8 alerts** when merging 72eb208ae724dfbc291218622f30c48505fa0136 into afdfe6879589d2c21aa73f84def1f53d9b8a3211 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3e269e05c2e79901f9d205c8dde2c16a7aeda9a5)\n\n**new alerts:**\n\n* 8 for Unused import"
] | 2020-11-11T21:59:04 | 2020-11-19T04:44:25 | 2020-11-19T04:44:20 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1440",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1440",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1440.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1440.patch",
"merged_at": "2020-11-19T04:44:20"
} | This PR adds the `.register_bert_model` and `.register_tokenizer` methods to `NLPModel` and lifts the `_setup_tokenizer` method to `NLPModel`.
For HuggingFace and Megatron BERT models the BERT config will be written to `pretrained_model_name_encoder_config.json` and added to the `.nemo` archive.
For tokenizers, the vocab file and tokenizer model will be added to the `.nemo` archive. If no vocab file is provided, and using HuggingFace AutoTokenizer, then the `vocab_dict` will extracted from the tokenizer object and the `vocab_file` will be created from the `vocab_dict` and both saved to the `.nemo` archive.
This gives Jarvis the necessary information to generate a TRT Engine from the PyTorch checkpoint.
Calling `.register_bert_model` from PTL `.setup` hook so that all models subclassed from `NLPModel` will register their bert model (if possible). | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1440/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1440/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1439 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1439/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1439/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1439/events | https://github.com/NVIDIA/NeMo/pull/1439 | 740,513,184 | MDExOlB1bGxSZXF1ZXN0NTE4OTg5MTM4 | 1,439 | vad outputs to manifest script for speaker diarization | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-11T06:45:04 | 2020-11-11T18:32:09 | 2020-11-11T18:32:01 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1439",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1439",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1439.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1439.patch",
"merged_at": "2020-11-11T18:32:01"
} | Signed-off-by: nithinraok <[email protected]> | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1439/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1439/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1438 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1438/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1438/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1438/events | https://github.com/NVIDIA/NeMo/pull/1438 | 740,489,321 | MDExOlB1bGxSZXF1ZXN0NTE4OTY4NzYw | 1,438 | EFF save/restore update | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-11T06:01:16 | 2021-01-22T15:29:32 | 2020-11-12T21:42:44 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1438",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1438",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1438.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1438.patch",
"merged_at": "2020-11-12T21:42:44"
} | Switching to EFF NeMoCookbook. | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1438/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1438/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1437 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1437/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1437/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1437/events | https://github.com/NVIDIA/NeMo/pull/1437 | 740,236,195 | MDExOlB1bGxSZXF1ZXN0NTE4NzU4ODE5 | 1,437 | Update NeMo core for melgan | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-10T20:55:23 | 2020-11-25T16:50:14 | 2020-11-11T06:43:34 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1437",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1437",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1437.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1437.patch",
"merged_at": "2020-11-11T06:43:34"
} | Split off updates to nemo core that are required for #1418
Changes:
- Remove empty *_step() methods from ModelPT
- Treat None inside of ElementType's parameters as a Void that matches anything
- Use case being removing the default 16kHz sample_rate from AudioSignal and replacing it with None. This allows a module with type AudioSignal() to match ANY AudioSignal regardless of sample_rate
Minor Changes:
- Add Singleton NeuralType
- Update wrap_training_step such it is doesn't require a dictionary to be passed
- Other cosmetic changes | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1437/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1437/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1436 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1436/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1436/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1436/events | https://github.com/NVIDIA/NeMo/issues/1436 | 739,965,249 | MDU6SXNzdWU3Mzk5NjUyNDk= | 1,436 | [Question] ASR Correction output | {
"login": "amejri",
"id": 45438779,
"node_id": "MDQ6VXNlcjQ1NDM4Nzc5",
"avatar_url": "https://avatars.githubusercontent.com/u/45438779?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/amejri",
"html_url": "https://github.com/amejri",
"followers_url": "https://api.github.com/users/amejri/followers",
"following_url": "https://api.github.com/users/amejri/following{/other_user}",
"gists_url": "https://api.github.com/users/amejri/gists{/gist_id}",
"starred_url": "https://api.github.com/users/amejri/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/amejri/subscriptions",
"organizations_url": "https://api.github.com/users/amejri/orgs",
"repos_url": "https://api.github.com/users/amejri/repos",
"events_url": "https://api.github.com/users/amejri/events{/privacy}",
"received_events_url": "https://api.github.com/users/amejri/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"The best bet it to train on more data.\r\nIf you do not have more data, then try fine-tuning from pretrained model (even from another language) by re-using acoustic encoder.\r\nAnd, of course, you will need to play/experiment with hyperparameters"
] | 2020-11-10T14:37:39 | 2020-11-11T00:19:47 | 2020-11-11T00:19:47 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello,
I'm trying to train ASR models for different languages. The problem is that the output contains many mistakes. Do you have any idea to correct them.
Thanks in advance,
Sincerely. | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1436/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1436/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1435 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1435/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1435/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1435/events | https://github.com/NVIDIA/NeMo/issues/1435 | 739,773,675 | MDU6SXNzdWU3Mzk3NzM2NzU= | 1,435 | How to export model from the checkpoint and then load the same model for prediction? | {
"login": "Khushali22",
"id": 10240131,
"node_id": "MDQ6VXNlcjEwMjQwMTMx",
"avatar_url": "https://avatars.githubusercontent.com/u/10240131?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Khushali22",
"html_url": "https://github.com/Khushali22",
"followers_url": "https://api.github.com/users/Khushali22/followers",
"following_url": "https://api.github.com/users/Khushali22/following{/other_user}",
"gists_url": "https://api.github.com/users/Khushali22/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Khushali22/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Khushali22/subscriptions",
"organizations_url": "https://api.github.com/users/Khushali22/orgs",
"repos_url": "https://api.github.com/users/Khushali22/repos",
"events_url": "https://api.github.com/users/Khushali22/events{/privacy}",
"received_events_url": "https://api.github.com/users/Khushali22/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"You should be able to:\r\n```\r\nmymodel.save_to(save_path=\"mymodel.nemo\")\r\n```\r\nthen (possibly in another python script)\r\n```\r\nmymodel=IntentSlotClassificationModel.restore_from(restore_path=\"mymodel.nemo\")\r\n```\r\nNote that ```restore_from``` is a class method which returns class instance. Are you calling it from existing instance instead when you say it is not working?\r\n",
"This example from our code shows also how to restore this model and use it for Inference.\r\nhttps://github.com/NVIDIA/NeMo/blob/main/examples/nlp/intent_slot_classification/intent_slot_classification.py",
"Yes, I have tried to call from existing instance as well as in another script as well. The error which i get is like these:-\r\n\r\nFileNotFoundError: Make sure that your data follows the standard format supported by JointIntentSlotDataset. Your data must contain dict.intents.csv and dict.slots.csv.\r\n\r\nI am assuming in this \"intent_slot_classification_descriptor.py \" file, this things has been defined in _init_ method constructor, but if i am restoring then i need to place intents & slots csv in some \"DATA_DIR\" folder? My question is if i am restoring model, really data needs to be there while restoring models?\r\n\r\nMany Thanks!",
"Basically the model currently expects to get Dictionary of Intents and Slots as files from outside, from \"data_dir\" folder. The reason for that is that the model is self sufficient but to interpret the outputs (numbers) of the classified Intents and Slots coming from the top layer of the model you still need to have a dictionary of appropriate labels. Please let us know if it creates any difficulties for you?",
"Many Thanks for your response. I am still not able to get my problem solved.\r\n\r\nLet me define the steps which i am following:-\r\n\r\nRun in google colab session-1:-\r\n\r\n1) Preprocess data into Nemo-format & create files (dict.intents.csv, dict.slots.csv, train.tsv, test.tsv, train_slots.tsv, test_slots.tsv)\r\n2) Fetch config yaml file & update parameters \r\n3) Do training\r\n4) model.save_to('model_education.nemo') # save the model at some drive location\r\n5) Evaluate from the checkpoint saved by model training:-\r\n`# extract the path of the best checkpoint from the training, you may update it to any other saved checkpoint file\r\ncheckpoint_path = trainer.checkpoint_callback.best_model_path\r\neval_model = nemo_nlp.models.IntentSlotClassificationModel.load_from_checkpoint(checkpoint_path=checkpoint_path)`\r\n6) Predict queries:-\r\n`queries = [\r\n 'hello',\r\n 'I would like to enroll for deep learning course',\r\n 'I enrolled and paid for DLNLP course which starts from October 10th',\r\n 'pycharm connection error'\r\n]\r\npred_intents, pred_slots = eval_model.predict_from_examples(queries)\r\n`\r\n\r\nI am able to predict from the above method as the training is in the same session.\r\n\r\nNow my `google colab session-1` is disconnected.\r\n\r\nAnother google colab session-2:-\r\n\r\n1) Import libraries\r\n2) Fetch config file\r\n3) Update DATA_DIR & NEMO_DIR Folder path\r\n4) Dict.intents.csv & dict.slots.csv files are placed at DATA_DIR Folder.\r\n5) Load the model\r\n\r\n`mymodel=nemo_nlp.models.IntentSlotClassificationModel.restore_from(restore_path=\"/content/drive/My Drive/Colab Notebooks/mymodel_edu.nemo\")`\r\n\r\nThen error comes:-\r\nFileNotFoundError: Make sure that your data follows the standard format supported by JointIntentSlotDataset. Your data must contain dict.intents.csv and dict.slots.csv.\r\n\r\nIt will be very helpful if you can help me where i made mistake or any other documentation i can refer which tells this kind of scenario?\r\n\r\nThanks.\r\n",
"hello @Khushali22 \r\nerror says file not found and from your explanation I'am seeing that you saved the nemo model with name as _'model_education.nemo')_ in step 4 and restoring it with name 'mymodel_edu.nemo' ? \r\n\r\n"
] | 2020-11-10T10:13:24 | 2021-08-17T00:56:20 | 2021-08-17T00:56:20 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi,
I have trained my own dataset using NLP Joint Intent & slot classification NeMo models. I am able to train & predict the data. However i am not getting how can i save the model from that checkpoint and then reload the same model & predict on example test queries. I can understand using experiment manager, path has been defined in yaml file for saving checkpoints, but i am unable to see that checkpoint file in that location(Running in google Colab).
My requirement is to save that model at my own defined location, then reload that model and i should able to predict from the reloaded model.
I have tried these things:-
1) save_to method (It's working. model has been saved)
2) restore_from() method (It's not working)
3) load_from_checkpoint (This is working, but i need to train always first, the i can load, But i dont want to train everytime. I want to just load model & predict from that trained model)
Need your help to understand the correct way of using saving reloading APIs of NeMo & why it is like not able to see that checkpoint in google Colab location & the same path i am able to reload from the checkpoint, but can't see that file.
**Environment overview (please complete the following information)**
- Environment location: Colab
- Method of NeMo install:
BRANCH = 'v1.0.0b2'
!python -m pip install git+https://github.com/NVIDIA/NeMo.git@$BRANCH#egg=nemo_toolkit[nlp]
Thank you!
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1435/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1435/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1434 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1434/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1434/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1434/events | https://github.com/NVIDIA/NeMo/pull/1434 | 739,701,884 | MDExOlB1bGxSZXF1ZXN0NTE4MzE2ODAx | 1,434 | Add support for Tarred dataset for Transformer LM | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging b7801269d9ee2ca213f5bd43e4302fd148ed66bc into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-fa359a48d5f95a10eb6e93d11f38e27ea636a1e6)\n\n**new alerts:**\n\n* 1 for Unused local variable\n* 1 for Unused import",
"This pull request **introduces 2 alerts** when merging 4433d5a86d59373b9e0250d55e11efeed56a221f into 5dbe245d04d4a062aaf42bd414ce72dfcdcbb092 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-66b8a18e76d721f10280ed97e01d9f333f971244)\n\n**new alerts:**\n\n* 1 for Unused local variable\n* 1 for Unused import",
"This pull request **introduces 2 alerts** when merging 49eb7b4238804bebe1bb4fee6a83f171160de066 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-6a3ee9a8d9291ff992ba1a3e3e39e4465f4aaf94)\n\n**new alerts:**\n\n* 1 for Unused local variable\n* 1 for Unused import",
"This pull request **introduces 2 alerts** when merging 325a9e15542900671730a0a78de7043244a315cd into 48bd21309145a6772fca7375d3ced377033dd3cf - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-acd1f51a693f4dccee1ce4e83bb806da8c454bdd)\n\n**new alerts:**\n\n* 1 for Unused local variable\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 61acf63635b6b437345a0123f4d3bcd70ffbd893 into 5b6dd3850129898be47cf0d65587897ec45a5b59 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ee293a8e9cfdfb6b110aed089b11822e23e2cab9)\n\n**new alerts:**\n\n* 1 for Unused local variable",
"@AlexGrinch The tokenizer can be either cased or uncased - for example LS normalized text corpus is all lower case chars, but other datasets may be cased. When training a LM with a custom vocabulary, it is worthwhile to have an option to lowercase the text prior to tokenization if required, so that additional processing is not required inside the tarred dataset. ",
"@titu1994 Please correct me if I am wrong. What I understood from the code is that in the process of creating tarred dataset, we take all text, transform it into token_ids via _tokenizer.text_to_ids(text)_, and save this numbers into tarred numpy array. If our tokenizer works with lower case text, text.lower() is called somewhere inside _tokenizer.text_to_ids(text)_. It seems that there is no need to do it in advance as we call _tokenizer.text_to_ids(text)_ either way during dataset creation.",
"For ASR datasets (as an example) - Librispeech LM corpus - the tokenizer is a bert base cased. The language corpus itself is in all Upper case, but the tokenizer is built to support only lower case. Since it is a cased tokenizer, it will convert all text to [UNK] instead of lowering it. \n\nSince this script directly uses a prebuilt tokenizer vocabulary therefore it is necessary to provide the logic to explicitly lowercase via the flag for compatibility between the text corpus (cases or uncased), the tokenizer vocabulary (cased or uncased) and the tokenizer construction (cased or uncased).",
"Okay, I think we can merge this PR after it passes all checks.",
"Sounds good, I'll fix up the error "
] | 2020-11-10T08:33:37 | 2020-11-18T21:08:15 | 2020-11-18T21:08:12 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1434",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1434",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1434.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1434.patch",
"merged_at": "2020-11-18T21:08:12"
} | # Changelog
- Adds tarred dataset support for transformer based Language models
- Pre-tokenizes and chunks the dataset for more efficient external dataset handling
- Adds utility script to construct a tarred tokenized LM dataset
- Adds `cache_ids` flag support for transformer based language models | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1434/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1434/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1433 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1433/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1433/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1433/events | https://github.com/NVIDIA/NeMo/issues/1433 | 739,684,641 | MDU6SXNzdWU3Mzk2ODQ2NDE= | 1,433 | How can i the change the provided classes on the mapping into my own [Token_Classification_Named_Entity_Recognition] | {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"If your labels are not a subset of the provided labels, it's recommended to use [this script](https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/token_classification/token_classification.py). Note, you `shouldn't` use a pre-trained TokenClassificationModel, instead start fine-tuning with the original BERT model. When you provide a path to your data directory, the text to int labels mapping will be created automatically based on the training data.\r\n\r\nThe data directory should follow the format described in the script or in the [tutorial](https://github.com/NVIDIA/NeMo/blob/main/tutorials/nlp/Token_Classification_Named_Entity_Recognition.ipynb). ",
"I now have loaded the files into `DATA_DIR = \"DATA_DIR\"` but I still get the same error though it's a different tag that is caught (B-GPU above but now B-Price)\r\n\r\nI also stopped the pre-trained TokenClassificationModel and ran the original BERT model\r\n\r\n",
"Could you please try following the `Model Training From Scratch` section of the [tutorial](https://github.com/NVIDIA/NeMo/blob/main/tutorials/nlp/Token_Classification_Named_Entity_Recognition.ipynb) and initialize your model with `model = nemo_nlp.models.TokenClassificationModel(cfg=config.model, trainer=trainer)`?\r\nThis should create labels mapping from your training data.\r\n\r\nIf you're running locally, it might be easier to use the training script with the default config: \r\n`python examples/nlp/token_classification/token_classification.py model.dataset.data_dir=<PATH_TO_DATA_DIR>`\r\n\r\nIf the issue persists, please add the output of the model initialization call here.",
"I got it to work now thank you!"
] | 2020-11-10T08:06:29 | 2020-11-11T04:09:57 | 2020-11-11T04:09:57 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi, I am using the Token_Classification_Named_Entity_Recognition tutorial for a project.
By default, the dataset provided these labels:
LOC = Geographical Entity
ORG = Organization
PER = Person
GPE = Geopolitical Entity
TIME = Time indicator
ART = Artifact
EVE = Event
NAT = Natural Phenomenon
How can I change the classes in the provided mapping such that it can use my labels and text file instead?
(My labels include: label_names = ['O', 'B-Product', 'B-Processor', 'B-WindowVersion','B-GPU','B-Display','B-MemoryStorage','B-Storage','B-Weight','B-Warranty','B-Price'] )
Attached is the error I received.
| {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1433/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1433/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1432 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1432/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1432/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1432/events | https://github.com/NVIDIA/NeMo/pull/1432 | 739,656,563 | MDExOlB1bGxSZXF1ZXN0NTE4Mjc5NjM2 | 1,432 | export() returns list of file names created | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"> This looks fine to me. But I am wondering where it will be used?\r\n\r\nIn EFF: ONNXCookbook :)"
] | 2020-11-10T07:18:12 | 2020-11-13T19:47:41 | 2020-11-13T19:28:04 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1432",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1432",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1432.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1432.patch",
"merged_at": "2020-11-13T19:28:04"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1432/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1432/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1431 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1431/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1431/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1431/events | https://github.com/NVIDIA/NeMo/pull/1431 | 739,457,609 | MDExOlB1bGxSZXF1ZXN0NTE4MTE3NjQ5 | 1,431 | Update tarred dataset conversion script | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-10T00:00:37 | 2020-11-10T18:48:17 | 2020-11-10T01:08:29 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1431",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1431",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1431.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1431.patch",
"merged_at": "2020-11-10T01:08:29"
} | Bugfix for when max duration is set to None
Add support for min duration filtering | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1431/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1431/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1430 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1430/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1430/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1430/events | https://github.com/NVIDIA/NeMo/pull/1430 | 739,456,064 | MDExOlB1bGxSZXF1ZXN0NTE4MTE2Mzc4 | 1,430 | Update tarred dataset conversion script | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 6 alerts** when merging 57bbdd1984e6529cb18f138488bad775e10a19b1 into c0fedd91e316a4ba95ac6b14330ff92844c56097 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-fc89a4569b49ec7de84b75ce25de8500ede41b9d)\n\n**new alerts:**\n\n* 4 for Unused import\n* 2 for Syntax error"
] | 2020-11-09T23:56:54 | 2020-11-10T00:10:14 | 2020-11-09T23:57:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1430",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1430",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1430.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1430.patch",
"merged_at": null
} | Bugfix for if maximum duration set to None
Added minimum duration filter | {
"login": "redoctopus",
"id": 4268876,
"node_id": "MDQ6VXNlcjQyNjg4NzY=",
"avatar_url": "https://avatars.githubusercontent.com/u/4268876?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/redoctopus",
"html_url": "https://github.com/redoctopus",
"followers_url": "https://api.github.com/users/redoctopus/followers",
"following_url": "https://api.github.com/users/redoctopus/following{/other_user}",
"gists_url": "https://api.github.com/users/redoctopus/gists{/gist_id}",
"starred_url": "https://api.github.com/users/redoctopus/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/redoctopus/subscriptions",
"organizations_url": "https://api.github.com/users/redoctopus/orgs",
"repos_url": "https://api.github.com/users/redoctopus/repos",
"events_url": "https://api.github.com/users/redoctopus/events{/privacy}",
"received_events_url": "https://api.github.com/users/redoctopus/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1430/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1430/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1429 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1429/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1429/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1429/events | https://github.com/NVIDIA/NeMo/pull/1429 | 738,840,234 | MDExOlB1bGxSZXF1ZXN0NTE3NjA2OTkw | 1,429 | Add RIR perturbation example. | {
"login": "Honghe",
"id": 1092722,
"node_id": "MDQ6VXNlcjEwOTI3MjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1092722?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Honghe",
"html_url": "https://github.com/Honghe",
"followers_url": "https://api.github.com/users/Honghe/followers",
"following_url": "https://api.github.com/users/Honghe/following{/other_user}",
"gists_url": "https://api.github.com/users/Honghe/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Honghe/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Honghe/subscriptions",
"organizations_url": "https://api.github.com/users/Honghe/orgs",
"repos_url": "https://api.github.com/users/Honghe/repos",
"events_url": "https://api.github.com/users/Honghe/events{/privacy}",
"received_events_url": "https://api.github.com/users/Honghe/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-09T09:09:11 | 2020-11-13T00:05:07 | 2020-11-13T00:05:07 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1429",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1429",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1429.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1429.patch",
"merged_at": "2020-11-13T00:05:07"
} | Add RIR(Room impulse Reverbation) perturbation example in the tutorial. | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1429/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1429/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1428 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1428/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1428/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1428/events | https://github.com/NVIDIA/NeMo/issues/1428 | 738,412,606 | MDU6SXNzdWU3Mzg0MTI2MDY= | 1,428 | [Question] Difference between `blank` and `space` character in ASR quartznet? | {
"login": "ORlGlN",
"id": 35373089,
"node_id": "MDQ6VXNlcjM1MzczMDg5",
"avatar_url": "https://avatars.githubusercontent.com/u/35373089?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ORlGlN",
"html_url": "https://github.com/ORlGlN",
"followers_url": "https://api.github.com/users/ORlGlN/followers",
"following_url": "https://api.github.com/users/ORlGlN/following{/other_user}",
"gists_url": "https://api.github.com/users/ORlGlN/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ORlGlN/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ORlGlN/subscriptions",
"organizations_url": "https://api.github.com/users/ORlGlN/orgs",
"repos_url": "https://api.github.com/users/ORlGlN/repos",
"events_url": "https://api.github.com/users/ORlGlN/events{/privacy}",
"received_events_url": "https://api.github.com/users/ORlGlN/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"`space` character is a regular delimiter between words [` `]. Usually, it is also returned for silent segments. \r\n`blank` symbol is a special extra character in CTC modeling that allows representing double characters, for example. It is removed from final decoded transcript. More details could be found in the following great paper: https://distill.pub/2017/ctc/ .",
"That's very helpful, thanks "
] | 2020-11-08T07:41:05 | 2020-11-10T19:24:15 | 2020-11-10T19:24:15 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | How to differentiate the meaning of these two characters? Which one represents silence, gap for next word or the end of word? Thanks in advance :) | {
"login": "ORlGlN",
"id": 35373089,
"node_id": "MDQ6VXNlcjM1MzczMDg5",
"avatar_url": "https://avatars.githubusercontent.com/u/35373089?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ORlGlN",
"html_url": "https://github.com/ORlGlN",
"followers_url": "https://api.github.com/users/ORlGlN/followers",
"following_url": "https://api.github.com/users/ORlGlN/following{/other_user}",
"gists_url": "https://api.github.com/users/ORlGlN/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ORlGlN/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ORlGlN/subscriptions",
"organizations_url": "https://api.github.com/users/ORlGlN/orgs",
"repos_url": "https://api.github.com/users/ORlGlN/repos",
"events_url": "https://api.github.com/users/ORlGlN/events{/privacy}",
"received_events_url": "https://api.github.com/users/ORlGlN/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1428/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1428/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1427 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1427/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1427/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1427/events | https://github.com/NVIDIA/NeMo/issues/1427 | 738,174,179 | MDU6SXNzdWU3MzgxNzQxNzk= | 1,427 | [Question] How may i save the model to my own file in colab? | {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"You can use google.colab library to download files off of google colab - https://stackoverflow.com/questions/48774285/how-to-download-file-created-in-colaboratory-workspace.\r\n\r\nInstead of `example.txt` in that snippet, you can provide the full path to the checkpoint instead and it should (hopefully) download the file to your local drive.\r\n\r\n[EDIT]\r\nOr the more modern way of simply using the inbuilt file browser to the left on Google colab - following the second comment on the above link",
"Sorry I think what I meant to say is to save a copy of the model into training_session_folder on google colab as it by default does into nemo_experiments ",
"For that, you can change the experiment manager config - the `exp_dir` to be precise. Set the exp_dir to your training path and it should save model checkpoints there. ",
"\r\ni did this not sure if it is correct",
"Sure that's fine too, but you want a file name inside model_file, so it should be `/content/training_session_folder/ner_model.pt`",
"You can more easily just use model.save_to(model_file) and then use TokenClassificationModel.restore_from(model_file) for more easy saving and restoring of Nemo models. ",
"okay thank you so much for replying, really appreciate your help!!"
] | 2020-11-07T05:08:06 | 2020-11-10T06:11:27 | 2020-11-10T06:11:27 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | After running trainer.fit(model_from_scratch) i see that the models are actually saved in
" Epoch 0: val_loss reached 0.38735 (best 0.38735), saving model to /content/nemo_experiments/token_classification_model/2020-11-07_03-05-01/checkpoints/token_classification_model---val_loss=0.39-epoch=0.ckpt as top 3"
How can I access this model or save it into my own local file (training_session_folder)?
I need to use the model to pass in a text file and predict the labels
p.s. I'm a beginner pls help T.T
| {
"login": "181473h",
"id": 45226851,
"node_id": "MDQ6VXNlcjQ1MjI2ODUx",
"avatar_url": "https://avatars.githubusercontent.com/u/45226851?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/181473h",
"html_url": "https://github.com/181473h",
"followers_url": "https://api.github.com/users/181473h/followers",
"following_url": "https://api.github.com/users/181473h/following{/other_user}",
"gists_url": "https://api.github.com/users/181473h/gists{/gist_id}",
"starred_url": "https://api.github.com/users/181473h/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/181473h/subscriptions",
"organizations_url": "https://api.github.com/users/181473h/orgs",
"repos_url": "https://api.github.com/users/181473h/repos",
"events_url": "https://api.github.com/users/181473h/events{/privacy}",
"received_events_url": "https://api.github.com/users/181473h/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1427/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1427/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1426 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1426/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1426/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1426/events | https://github.com/NVIDIA/NeMo/pull/1426 | 738,113,841 | MDExOlB1bGxSZXF1ZXN0NTE3MDM4MDcy | 1,426 | Add support for replicated shard strategy for tarred datasets | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"As in remove the configs shard_strategy = scatter portion? We could drop them, but it reduces visibility. Practically, it should be fine, since the shard strategy is not used unless is_tarred is set \r\n\r\nEdit: I do agree that the configs are going quite big though. ",
"I was worried about it potentially causing some confusion, since it's not explicitly named in relation to the tarred dataset. How do you feel about a `null` option like for `tarred_audio_filepaths` when `is_tarred` is False?",
"Oh that's a good option. None by default switches to the current scatter strategy and otherwise the user sets it explicitly. I'll make the changes ",
"Another is to name the parameter `tarred_shard_strategy` ",
"Either sounds good to me",
"Thanks !"
] | 2020-11-07T00:16:58 | 2020-11-09T23:43:12 | 2020-11-09T23:42:55 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1426",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1426",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1426.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1426.patch",
"merged_at": "2020-11-09T23:42:55"
} | # Changelog
- Support `replicate` strategy for sharding of tarfiles under ddp for Tarred WebDataset.
- Add config line to all ASR models
Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1426/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1426/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1425 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1425/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1425/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1425/events | https://github.com/NVIDIA/NeMo/issues/1425 | 737,770,909 | MDU6SXNzdWU3Mzc3NzA5MDk= | 1,425 | [Question]Quantized QuartzNet | {
"login": "ABoltachev",
"id": 30128653,
"node_id": "MDQ6VXNlcjMwMTI4NjUz",
"avatar_url": "https://avatars.githubusercontent.com/u/30128653?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ABoltachev",
"html_url": "https://github.com/ABoltachev",
"followers_url": "https://api.github.com/users/ABoltachev/followers",
"following_url": "https://api.github.com/users/ABoltachev/following{/other_user}",
"gists_url": "https://api.github.com/users/ABoltachev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ABoltachev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ABoltachev/subscriptions",
"organizations_url": "https://api.github.com/users/ABoltachev/orgs",
"repos_url": "https://api.github.com/users/ABoltachev/repos",
"events_url": "https://api.github.com/users/ABoltachev/events{/privacy}",
"received_events_url": "https://api.github.com/users/ABoltachev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | null | [] | null | [
"I don't think they can help with this\r\nPytorch need to support Conv1d and BatchNorm1d in order for you to quantize it.\r\n\r\nThe only way you can do it is by converting Conv1d to Conv2d(same for BatchNorm) using [this](https://github.com/NVIDIA/NeMo/blob/c90e3aaee67a73dc1c2b0c0614766246e403e1c1/nemo/utils/export_utils.py#L53)\r\n\r\nThat code not working right now, you need to change some stuff to make it to work",
"we are actually working on adding quantization support, but it is not there yet.",
"I apologize, I was wrong with the PyTorch version, I used 1.4.0 but have now updated to 1.6.0\r\nThe problem with Conv1d is gone as it is supported in version 1.6.0. Also the issue with BatchNorm1d has been resolved (https://github.com/pytorch/pytorch/issues/41642#issuecomment-668115459). However, now, I have a problem with ReLU.\r\nFor some reason, not all ReLUs are quantized and, most likely because of this, at the trace stage in the line https://github.com/NVIDIA/NeMo/blob/v0.10.1/nemo/collections/asr/parts/jasper.py#L425 throws \"RuntimeError: _Map_base :: at\" exception\r\nIf necessary, I can send the full output of the architecture after quantization"
] | 2020-11-06T13:49:15 | 2021-06-19T04:54:39 | 2021-06-19T04:54:39 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am trying to quantize a QuartzNet 15x5 encoder like this:
```python
jasper_encoder = jasper_encoder.cpu()
jasper_encoder.eval()
jasper_encoder.qconfig = torch.quantization.get_default_qconfig('qnnpack')
torch.quantization.propagate_qconfig_(jasper_encoder)
jasper_encoder = torch.quantization.prepare(jasper_encoder, inplace=False)
jasper_encoder = torch.quantization.convert(jasper_encoder)
```
But, unfortunately, the encoder is not fully quantized, but only some RuLu. I also tried to quantize MaskedConv1d separately, but unfortunately even in this case the Conv1d layer was not quantized
This is what I get when I do print (jasper_encoder.encoder) after quantizing:
```
Sequential(
(0): JasperBlock(
(mconv): ModuleList(
(0): MaskedConv1d(
(conv): Conv1d(64, 64, kernel_size=[33], stride=[2], padding=(16,), dilation=[1], groups=64, bias=False)
)
(1): MaskedConv1d(
(conv): Conv1d(64, 256, kernel_size=(1,), stride=(1,), bias=False)
)
(2): BatchNorm1d(256, eps=0.001, momentum=0.1, affine=True, track_running_stats=True)
)
(mout): Sequential(
(0): QuantizedReLU()
(1): Dropout(p=0.0, inplace=False)
)
)
...
```
As you can see from this output, neither Conv1d nor BatchNorm1d were quantized. Could you help me with quantizing QuartzNet if it is possible to quantize this model
**Environment overview (please complete the following information)**
- Environment location: Bare-metal
- Method of NeMo install: pip install
**Environment details**
- OS version: Ubuntu 18.04 LTS
- PyTorch version : 1.6.0
- Python version: Python 3.6
- NeMo version: 0.10.1
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1425/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1425/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1424 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1424/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1424/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1424/events | https://github.com/NVIDIA/NeMo/issues/1424 | 737,681,590 | MDU6SXNzdWU3Mzc2ODE1OTA= | 1,424 | VAD offline inference | {
"login": "2Bye",
"id": 45552093,
"node_id": "MDQ6VXNlcjQ1NTUyMDkz",
"avatar_url": "https://avatars.githubusercontent.com/u/45552093?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/2Bye",
"html_url": "https://github.com/2Bye",
"followers_url": "https://api.github.com/users/2Bye/followers",
"following_url": "https://api.github.com/users/2Bye/following{/other_user}",
"gists_url": "https://api.github.com/users/2Bye/gists{/gist_id}",
"starred_url": "https://api.github.com/users/2Bye/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/2Bye/subscriptions",
"organizations_url": "https://api.github.com/users/2Bye/orgs",
"repos_url": "https://api.github.com/users/2Bye/repos",
"events_url": "https://api.github.com/users/2Bye/events{/privacy}",
"received_events_url": "https://api.github.com/users/2Bye/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"Hello @4nton-P \r\n\r\n- I just want to feed a track into the model and get a predicted time. To do this, I'll just rewrite the callback function, but how to get around the problem with the lack of devices?\r\nIt is designed for the demonstration in the tutorial you don't have to use it to do inference for VAD.\r\n\r\nWe're updating the inference pipeline here`https://github.com/NVIDIA/NeMo/pull/1384` and will be updated and merged soon. It will help you to perform **AUDIO -> MANIFEST -> VAD OUTPUT TXT -> VAD RTTM-like TABLE**\r\n\r\nI will let you know once it merged. Hope it helps. \r\n",
"> Hello @4nton-P\r\n> \r\n> * I just want to feed a track into the model and get a predicted time. To do this, I'll just rewrite the callback function, but how to get around the problem with the lack of devices?\r\n> It is designed for the demonstration in the tutorial you don't have to use it to do inference for VAD.\r\n> \r\n> We're updating the inference pipeline here`https://github.com/NVIDIA/NeMo/pull/1384` and will be updated and merged soon. It will help you to perform **AUDIO -> MANIFEST -> VAD OUTPUT TXT -> VAD RTTM-like TABLE**\r\n> \r\n> I will let you know once it merged. Hope it helps.\r\n\r\nThanks, I'll wait",
"@4nton-P The PR is merged. Feel free to play with the scripts. Questions and feedbacks are very welcome!\r\nI'm closing this issue. Please reopen or open a new one with questions. Thanks!"
] | 2020-11-06T11:19:57 | 2020-11-11T21:44:08 | 2020-11-11T21:44:08 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello!
I ran into the VAD Inference problem. I cannot use offline inference if I do not have an output or input of a device, because the pyaudio requires a device. Are there any solutions?
I just want to feed a track into the model and get a predicted time. To do this, I'll just rewrite the callback function, but how to get around the problem with the lack of devices? | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1424/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1424/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1423 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1423/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1423/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1423/events | https://github.com/NVIDIA/NeMo/pull/1423 | 737,386,853 | MDExOlB1bGxSZXF1ZXN0NTE2NDM3ODI1 | 1,423 | Fixed cd command syntax in notebooks | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-06T01:32:59 | 2020-11-06T06:38:45 | 2020-11-06T06:38:44 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1423",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1423",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1423.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1423.patch",
"merged_at": "2020-11-06T06:38:44"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1423/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1423/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1422 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1422/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1422/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1422/events | https://github.com/NVIDIA/NeMo/pull/1422 | 737,335,054 | MDExOlB1bGxSZXF1ZXN0NTE2Mzk2MDA4 | 1,422 | Fixes perplexity computation (issue https://github.com/NVIDIA/NeMo/issues/1289) | {
"login": "PeganovAnton",
"id": 10737305,
"node_id": "MDQ6VXNlcjEwNzM3MzA1",
"avatar_url": "https://avatars.githubusercontent.com/u/10737305?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/PeganovAnton",
"html_url": "https://github.com/PeganovAnton",
"followers_url": "https://api.github.com/users/PeganovAnton/followers",
"following_url": "https://api.github.com/users/PeganovAnton/following{/other_user}",
"gists_url": "https://api.github.com/users/PeganovAnton/gists{/gist_id}",
"starred_url": "https://api.github.com/users/PeganovAnton/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/PeganovAnton/subscriptions",
"organizations_url": "https://api.github.com/users/PeganovAnton/orgs",
"repos_url": "https://api.github.com/users/PeganovAnton/repos",
"events_url": "https://api.github.com/users/PeganovAnton/events{/privacy}",
"received_events_url": "https://api.github.com/users/PeganovAnton/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "PeganovAnton",
"id": 10737305,
"node_id": "MDQ6VXNlcjEwNzM3MzA1",
"avatar_url": "https://avatars.githubusercontent.com/u/10737305?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/PeganovAnton",
"html_url": "https://github.com/PeganovAnton",
"followers_url": "https://api.github.com/users/PeganovAnton/followers",
"following_url": "https://api.github.com/users/PeganovAnton/following{/other_user}",
"gists_url": "https://api.github.com/users/PeganovAnton/gists{/gist_id}",
"starred_url": "https://api.github.com/users/PeganovAnton/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/PeganovAnton/subscriptions",
"organizations_url": "https://api.github.com/users/PeganovAnton/orgs",
"repos_url": "https://api.github.com/users/PeganovAnton/repos",
"events_url": "https://api.github.com/users/PeganovAnton/events{/privacy}",
"received_events_url": "https://api.github.com/users/PeganovAnton/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "PeganovAnton",
"id": 10737305,
"node_id": "MDQ6VXNlcjEwNzM3MzA1",
"avatar_url": "https://avatars.githubusercontent.com/u/10737305?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/PeganovAnton",
"html_url": "https://github.com/PeganovAnton",
"followers_url": "https://api.github.com/users/PeganovAnton/followers",
"following_url": "https://api.github.com/users/PeganovAnton/following{/other_user}",
"gists_url": "https://api.github.com/users/PeganovAnton/gists{/gist_id}",
"starred_url": "https://api.github.com/users/PeganovAnton/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/PeganovAnton/subscriptions",
"organizations_url": "https://api.github.com/users/PeganovAnton/orgs",
"repos_url": "https://api.github.com/users/PeganovAnton/repos",
"events_url": "https://api.github.com/users/PeganovAnton/events{/privacy}",
"received_events_url": "https://api.github.com/users/PeganovAnton/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"This pull request **fixes 1 alert** when merging eb11ff4b16f70648862ee2e85a4912101d10c1d6 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-f332123bfafc9c57f8b57f0fc73bcbf047aea797)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 89bc627ed6df26dd20471ee7a848de79964d6a84 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-aca3524081cb9dd1c26ea5a0f50b3bfd93848444)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging fbc26fc722bda9ff51a5c5ce44b8b59dc5c0e4f6 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-88d367f4a639ea975e3a703a26cc799f11241726)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging f5b54e58748917389b1792c99bea5be131236709 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-88c4f53ef90666cccb4e8d4dfc2d679324a9cb73)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 0703eba7f634fd3423b0b0d30272028077a63ec0 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-24e865175ba9a1ebfd8b03c56148a767751006d4)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging d0c1a2e10d8524583a92aba4b6fd8e5e3b631ffe into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-c97c4bbc694eeee169f495a6cd9222885c188651)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 40e0096fcf99f77dea642bf885dbda79956927bf into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8955cc980ad0768de00b3dcfab9b90294ca9bc8a)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging bfd10487c05c6d22c49c842e5791d2806b098d23 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-bf54d165f39cdd3449a151f459152c3e482f24bc)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"@PeganovAnton Could you add the Perplexity metric to the Transformer based language model class in NeMo ? ",
"This pull request **fixes 1 alert** when merging f2441874c6b42c078e42e3ab16ca3e04a0fb7255 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-2a522744379d24a0f18b1737749d2b04e7ad0e42)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 147aa9d37029b43cfb3be7a9d0c43655448a0219 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-310184ee2f5eda2471ca8fdeb3289a3a75529a5b)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 830b46987f2d3d40462f44c78f9c8ccb5b4cddf0 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ec9a01906d2dd5159d6321ad9bfeab82823d5c8e)\n\n**fixed alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 1 alert** when merging 07414ccd56c07724b1c16b59800cb72ebbaca0b7 into 075472199a3b2b65530e232c6ece4ce6ef92bf28 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-2199a3ee791caa50fd1a5dad77f2dd2d5f57904d)\n\n**fixed alerts:**\n\n* 1 for Unused import"
] | 2020-11-05T23:16:36 | 2021-01-22T15:29:42 | 2020-11-11T07:56:38 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1422",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1422",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1422.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1422.patch",
"merged_at": "2020-11-11T07:56:38"
} | * Add perplexity draft
Signed-off-by: PeganovAnton <[email protected]>
* Add perplexity metric
Signed-off-by: PeganovAnton <[email protected]>
* Add docstrings and copyright, fix comments
Signed-off-by: PeganovAnton <[email protected]>
* Fix code style
Signed-off-by: PeganovAnton <[email protected]>
* Fix code style
Signed-off-by: PeganovAnton <[email protected]>
* Improve docstring
Signed-off-by: PeganovAnton <[email protected]>
* Fix docstring hrefs
Signed-off-by: PeganovAnton <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1422/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1422/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1421 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1421/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1421/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1421/events | https://github.com/NVIDIA/NeMo/pull/1421 | 737,326,366 | MDExOlB1bGxSZXF1ZXN0NTE2Mzg4NjM5 | 1,421 | Fix perplexity | {
"login": "PeganovAnton",
"id": 10737305,
"node_id": "MDQ6VXNlcjEwNzM3MzA1",
"avatar_url": "https://avatars.githubusercontent.com/u/10737305?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/PeganovAnton",
"html_url": "https://github.com/PeganovAnton",
"followers_url": "https://api.github.com/users/PeganovAnton/followers",
"following_url": "https://api.github.com/users/PeganovAnton/following{/other_user}",
"gists_url": "https://api.github.com/users/PeganovAnton/gists{/gist_id}",
"starred_url": "https://api.github.com/users/PeganovAnton/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/PeganovAnton/subscriptions",
"organizations_url": "https://api.github.com/users/PeganovAnton/orgs",
"repos_url": "https://api.github.com/users/PeganovAnton/repos",
"events_url": "https://api.github.com/users/PeganovAnton/events{/privacy}",
"received_events_url": "https://api.github.com/users/PeganovAnton/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **fixes 1 alert** when merging 95dfdd4e791577dfdcef46bb40b29ea6fc654767 into 1ac828df423fbcec1b34c650b3a20266bb133dde - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-49f4b5e0cd5f2f08dc1ab6bb582037306b47e847)\n\n**fixed alerts:**\n\n* 1 for Unused import"
] | 2020-11-05T22:57:10 | 2021-07-20T18:02:54 | 2020-11-05T23:12:39 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1421",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1421",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1421.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1421.patch",
"merged_at": "2020-11-05T23:12:39"
} | Fix for issue https://github.com/NVIDIA/NeMo/issues/1289 . Move code to the branch in the original repo for easier collaboration with @yzhang123 | {
"login": "PeganovAnton",
"id": 10737305,
"node_id": "MDQ6VXNlcjEwNzM3MzA1",
"avatar_url": "https://avatars.githubusercontent.com/u/10737305?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/PeganovAnton",
"html_url": "https://github.com/PeganovAnton",
"followers_url": "https://api.github.com/users/PeganovAnton/followers",
"following_url": "https://api.github.com/users/PeganovAnton/following{/other_user}",
"gists_url": "https://api.github.com/users/PeganovAnton/gists{/gist_id}",
"starred_url": "https://api.github.com/users/PeganovAnton/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/PeganovAnton/subscriptions",
"organizations_url": "https://api.github.com/users/PeganovAnton/orgs",
"repos_url": "https://api.github.com/users/PeganovAnton/repos",
"events_url": "https://api.github.com/users/PeganovAnton/events{/privacy}",
"received_events_url": "https://api.github.com/users/PeganovAnton/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1421/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1421/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1420 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1420/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1420/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1420/events | https://github.com/NVIDIA/NeMo/pull/1420 | 737,313,993 | MDExOlB1bGxSZXF1ZXN0NTE2Mzc4MjA0 | 1,420 | Fixed params backward compatibility for BPE | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-05T22:30:02 | 2021-05-03T21:53:53 | 2020-11-05T23:06:38 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1420",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1420",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1420.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1420.patch",
"merged_at": "2020-11-05T23:06:38"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1420/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1420/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1419 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1419/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1419/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1419/events | https://github.com/NVIDIA/NeMo/pull/1419 | 737,260,600 | MDExOlB1bGxSZXF1ZXN0NTE2MzMzNDU1 | 1,419 | Fixed params backward compatibility | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Thanks, Vahid! 👍 "
] | 2020-11-05T20:55:40 | 2020-11-05T21:34:57 | 2020-11-05T21:34:40 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1419",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1419",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1419.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1419.patch",
"merged_at": "2020-11-05T21:34:39"
} | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1419/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1419/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1418 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1418/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1418/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1418/events | https://github.com/NVIDIA/NeMo/pull/1418 | 737,238,576 | MDExOlB1bGxSZXF1ZXN0NTE2MzE1MDYy | 1,418 | Add GAN-based TTS Vocoder | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 6 alerts** when merging 1a32266b9ae76f036aac21b58486d9e9c54ebe40 into 4de7bd8394fb7e809e04ee8bcadcf9a1711ef497 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-f4743e03b4be34541d11903c6d31e9d4f4d054b4)\n\n**new alerts:**\n\n* 3 for Unused import\n* 1 for Signature mismatch in overriding method\n* 1 for Unused local variable\n* 1 for Suspicious unused loop iteration variable",
"This pull request **introduces 3 alerts** when merging 31efecac501313a482045fa84f4c9de37b1c0a43 into 4de7bd8394fb7e809e04ee8bcadcf9a1711ef497 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-213f5a3610ecb1c635b1218b7044c6fe988b1630)\n\n**new alerts:**\n\n* 1 for Signature mismatch in overriding method\n* 1 for Unused import\n* 1 for Wrong number of arguments in a class instantiation",
"This pull request **introduces 1 alert** when merging 1d83b438b70c1b622a1fdb773462fc4dfda432d8 into 4de7bd8394fb7e809e04ee8bcadcf9a1711ef497 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-da57d6a36fa0609eaa3daebd4bebdf361ce395c7)\n\n**new alerts:**\n\n* 1 for Signature mismatch in overriding method",
"This pull request **introduces 1 alert** when merging ea45a80522c99426c3e70465af8bb5a087b52624 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8cbef5ba362dbc3ec63a86ec04acafd90e0fc464)\n\n**new alerts:**\n\n* 1 for Unused import"
] | 2020-11-05T20:18:23 | 2020-11-26T16:00:55 | 2020-11-26T16:00:51 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1418",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1418",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1418.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1418.patch",
"merged_at": "2020-11-26T16:00:51"
} | - Adds the fullband MelGAN model as described in https://arxiv.org/abs/2005.05106
- Remove empty *_step() from ModelPT
- Split UniGlowLoss into glow portion and STFTLoss portion
- Simply stft_exact_pad option in FilterBankFeatures
ToDos
- [x] ~Enable FP16~. No currently a priority.
- [x] Add NeMo Loss subclass
- [x] Update Docs
- [x] ~Update TTS tutorials~, can't until I upload the .nemo files | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1418/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1418/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1417 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1417/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1417/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1417/events | https://github.com/NVIDIA/NeMo/issues/1417 | 737,104,616 | MDU6SXNzdWU3MzcxMDQ2MTY= | 1,417 | ASR Post Processing with language model with new Nemo version | {
"login": "sl1msa1d",
"id": 73840928,
"node_id": "MDQ6VXNlcjczODQwOTI4",
"avatar_url": "https://avatars.githubusercontent.com/u/73840928?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sl1msa1d",
"html_url": "https://github.com/sl1msa1d",
"followers_url": "https://api.github.com/users/sl1msa1d/followers",
"following_url": "https://api.github.com/users/sl1msa1d/following{/other_user}",
"gists_url": "https://api.github.com/users/sl1msa1d/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sl1msa1d/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sl1msa1d/subscriptions",
"organizations_url": "https://api.github.com/users/sl1msa1d/orgs",
"repos_url": "https://api.github.com/users/sl1msa1d/repos",
"events_url": "https://api.github.com/users/sl1msa1d/events{/privacy}",
"received_events_url": "https://api.github.com/users/sl1msa1d/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "AlexGrinch",
"id": 8689095,
"node_id": "MDQ6VXNlcjg2ODkwOTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/8689095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AlexGrinch",
"html_url": "https://github.com/AlexGrinch",
"followers_url": "https://api.github.com/users/AlexGrinch/followers",
"following_url": "https://api.github.com/users/AlexGrinch/following{/other_user}",
"gists_url": "https://api.github.com/users/AlexGrinch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AlexGrinch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AlexGrinch/subscriptions",
"organizations_url": "https://api.github.com/users/AlexGrinch/orgs",
"repos_url": "https://api.github.com/users/AlexGrinch/repos",
"events_url": "https://api.github.com/users/AlexGrinch/events{/privacy}",
"received_events_url": "https://api.github.com/users/AlexGrinch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "AlexGrinch",
"id": 8689095,
"node_id": "MDQ6VXNlcjg2ODkwOTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/8689095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AlexGrinch",
"html_url": "https://github.com/AlexGrinch",
"followers_url": "https://api.github.com/users/AlexGrinch/followers",
"following_url": "https://api.github.com/users/AlexGrinch/following{/other_user}",
"gists_url": "https://api.github.com/users/AlexGrinch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AlexGrinch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AlexGrinch/subscriptions",
"organizations_url": "https://api.github.com/users/AlexGrinch/orgs",
"repos_url": "https://api.github.com/users/AlexGrinch/repos",
"events_url": "https://api.github.com/users/AlexGrinch/events{/privacy}",
"received_events_url": "https://api.github.com/users/AlexGrinch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"sorry this feature have not been ported yet.",
"@okuchaiev is it wip or will take some time?",
"This is available in Nemo. Please refer to Nemo docs for ASR Language Modelling. "
] | 2020-11-05T16:58:12 | 2021-08-17T00:55:35 | 2021-08-17T00:55:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | How can we deploy ASR post processing on the new Nemo Version 1.0.0b1 ?
I found this [tutorial](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/v0.11.0/nlp/asr-improvement.html#asr-improvement) and this [page](https://github.com/NVIDIA/NeMo/tree/fa68d336cfefbbbb1849ff1b6ef454149f45234d/examples/nlp/asr_postprocessor) but they don't seem to work with the new Nemo Version
**Environment overview (please complete the following information)**
- Environment location: Bare-Metal with GPU / GCP
- Method of NeMo install: pip install
**Environment details**
- OS version: Ubuntu 20.04.1 LTS
- PyTorch version : 1.6
- Python version: Python 3.7.6
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1417/reactions",
"total_count": 4,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 2
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1417/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1416 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1416/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1416/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1416/events | https://github.com/NVIDIA/NeMo/issues/1416 | 736,911,804 | MDU6SXNzdWU3MzY5MTE4MDQ= | 1,416 | Segmentation fault (core dumped) when set create_tb_writer=True | {
"login": "caohongnga",
"id": 50764011,
"node_id": "MDQ6VXNlcjUwNzY0MDEx",
"avatar_url": "https://avatars.githubusercontent.com/u/50764011?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/caohongnga",
"html_url": "https://github.com/caohongnga",
"followers_url": "https://api.github.com/users/caohongnga/followers",
"following_url": "https://api.github.com/users/caohongnga/following{/other_user}",
"gists_url": "https://api.github.com/users/caohongnga/gists{/gist_id}",
"starred_url": "https://api.github.com/users/caohongnga/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/caohongnga/subscriptions",
"organizations_url": "https://api.github.com/users/caohongnga/orgs",
"repos_url": "https://api.github.com/users/caohongnga/repos",
"events_url": "https://api.github.com/users/caohongnga/events{/privacy}",
"received_events_url": "https://api.github.com/users/caohongnga/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1485815557,
"node_id": "MDU6TGFiZWwxNDg1ODE1NTU3",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": "Something isn't working"
}
] | closed | false | null | [] | null | [
"Thanks for the issue. Please upgrade your version of NeMo to 1.0.0b*.\r\n\r\nSee https://github.com/NVIDIA/NeMo/issues/1091 for a discussion on what has changed."
] | 2020-11-05T12:58:42 | 2020-12-04T00:46:20 | 2020-12-04T00:46:20 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Dear all,
I have following problem when I run speech2text file:
**Setup NeuralModuleFactory to control training nf = nemo.core.NeuralModuleFactory( create_tb_writer=True, ...)*
The system will raise an error:
################################################################################
### WARNING, path does not exist: KALDI_ROOT=/mnt/matylda5/iveselyk/Tools/kaldi-trunk
### (please add 'export KALDI_ROOT=<your_path>' in your $HOME/.profile)
### (or run as: KALDI_ROOT=<your_path> python <your_script>.py)
################################################################################
Segmentation fault (core dumped)
**Environment overview (please complete the following information)**
- Environment location: anaconda, nemo v0110
- Method of NeMo install: pip install nemo_toolkit[all]
**Environment details**
Otherwise, please provide:
- OS version: Ubuntu 16.04.5 LTS
- PyTorch version: pytorch 1.6.0
- Python version: Python 3.7.7
**Additional context**
CUDA Version: 10.1
GPU: GeForce GTX 1080
Please show me where was wrong!
Thank you very much!
Cao Hong Nga
| {
"login": "caohongnga",
"id": 50764011,
"node_id": "MDQ6VXNlcjUwNzY0MDEx",
"avatar_url": "https://avatars.githubusercontent.com/u/50764011?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/caohongnga",
"html_url": "https://github.com/caohongnga",
"followers_url": "https://api.github.com/users/caohongnga/followers",
"following_url": "https://api.github.com/users/caohongnga/following{/other_user}",
"gists_url": "https://api.github.com/users/caohongnga/gists{/gist_id}",
"starred_url": "https://api.github.com/users/caohongnga/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/caohongnga/subscriptions",
"organizations_url": "https://api.github.com/users/caohongnga/orgs",
"repos_url": "https://api.github.com/users/caohongnga/repos",
"events_url": "https://api.github.com/users/caohongnga/events{/privacy}",
"received_events_url": "https://api.github.com/users/caohongnga/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1416/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1416/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1415 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1415/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1415/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1415/events | https://github.com/NVIDIA/NeMo/pull/1415 | 736,530,749 | MDExOlB1bGxSZXF1ZXN0NTE1NzI4NzAx | 1,415 | Classical Transformer NMT architecture | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging 9adc613c98648adec02bcb31ce0ca116ddc499a5 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ee887455e4bef6a24010d44d6c8e730b052dac08)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 1d4dad90fe643619fb5cd2dec1b490f321872d28 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-9cb040e371cb44e97a548e8cf44073726f3ca009)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 69107446543481fbac8bcfeeee59c2086c3d759b into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-49c88be62644baf8a3a982738af75818c87cdc56)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 533dcded9d143eb9de47b2124a00a497dca09ac1 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-e5e03916ff220185b044649787a3425f6e3e6553)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 2eae2b12dd520b076742ec8fa22df42030b4f6b2 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-1bd96cd563cc994e82cf029dd3efe1a8fc449674)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 6dda1ee6e5e06a9345df712f101e1a2c4be075b8 into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ea603737d4a059249a0d50260685a464905c6fa9)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 1f6401ad174a5a2464b97d10c77aecf96a7019dc into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-cf06011c5f65d17024ad5872dd16b42e18233ea9)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 26096e6a86c770f2dcdeac43d9a1df12cc2d9dfc into d28114ae203558a0bd1bcc69e78aeb957ea9ed5e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-b13becc6b891a0d86697b7bbc85d27cfcabb2222)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 25a8d5ea3c8de5cbfbc06df4fd85a0c1b7496daa into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-33ff8544643c19627f981f179ba24de9210f1432)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 87fbbdcb8e8f5316e871322d6e05ad65d06ccf37 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-9d4a0e4e6714daf253f90e2aaf535aa70e778ba0)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 949a5c1e2eb10cae6b1a8bdc21cd2ab620f084f2 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-828c1d1a73362c8369e89cc7d9229df974c3978c)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging dd2c24ebc1a70455d5a9c154773b078a2debe1a3 into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-2ff5c3d474e961283ad6d47d5bd191b893534dbf)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 4795892f6c90bbbb077abf23f26364469b9843d1 into ffb4e36be574eda9fa35a14a65a3bfd372615ff0 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-d5e0a14245942730f6b2287fa3b19518d8431d0f)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 81264247a6326a79e5f3797eb4137ae70c9d9433 into ffb4e36be574eda9fa35a14a65a3bfd372615ff0 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3d5204067f9777903ebf9b950a20ab52246ad06c)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging ad8441d01e3036c1a413b1fc0fcf3b5940e70824 into ffb4e36be574eda9fa35a14a65a3bfd372615ff0 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-7ca2fbd033e4f39774d1c2677c1422287d08edb5)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 49f7c88988f2211f6bd714a6cf5b00df49c75e93 into ffb4e36be574eda9fa35a14a65a3bfd372615ff0 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-6e26822959a5445cfc83b20240ee429847eec1ba)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging 80206815e171e55e5ece20d66bba38313ee4fe40 into ffb4e36be574eda9fa35a14a65a3bfd372615ff0 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8e5d46345f6b1977ed746097340217135f619e79)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging c049297633811aec4df6ab6cbc89a1c814d047ec into 5dbe245d04d4a062aaf42bd414ce72dfcdcbb092 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-c11f73dab73291157a8b4cc8b9a157a492d5cdee)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 2 alerts** when merging 95c8dfeacf15e68ecf9fb93aa0367b3d7c041067 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-27c9552c866c6f1a1a0a95cba299f3727cbed363)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging cd781e52b1b21a3506b8b1f8987d8b7c004ac201 into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-afb731867b7070ec22dff1525b4ff1654e34fc4e)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging d3e8fdeb329b9e04496bef4a159b4acdf6696a4a into a711ca1ad2987797b1e80ac9b0ffd7479c0a4730 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8dea1c4a565e4b4ab206f057bd738a0ab5fd006d)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging f5e3a3b994308335caf495fe4b84783799a47954 into 5b6dd3850129898be47cf0d65587897ec45a5b59 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-15169ef3da1bb914e3bb276f8906b2123efb478e)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 2 alerts** when merging 854d8c1ab73f93dcb757678f3bfca0866941dec5 into 5b6dd3850129898be47cf0d65587897ec45a5b59 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-3ff9ec3184af0e552a3e5b5c6cbd6a2433d18bcf)\n\n**new alerts:**\n\n* 2 for Unused import",
"This pull request **introduces 3 alerts** when merging cfc84cebd98d92fa06089cdd42da685e60521d26 into 9ebc53426421edda35fd417f2a4c183daf2418f4 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-85eddb53ee5990245302fce388376ed68634704b)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging ddfae2e8988cd4d518e241daad1a75660735f2ba into 5572c82c0ba0c160eb940e94c2954886203abba6 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-46645bb0f5de2557e660b424fb63ff862cd1bfa1)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging 5cf41d07065cec230ecb4476faef3576df3e6642 into 6ae02c91eadcf0293d767597231ea7803aca8ce6 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ae24ca850b05d4566070b93d5ffa76c59d239845)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging 17968f8ef941d960bc28f646e1951b65026a6a14 into 70d1c87985714805506082feacfa3ca3ddcb4401 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-77fae6d506c64bfa0a37b204dfb13533e8895c4f)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging 631eaf9845d77708d6922dae5b5ac17805d19026 into 70d1c87985714805506082feacfa3ca3ddcb4401 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8dc029ff8ffa6c9964f21d41b37e1f839e7f98cf)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging 96ccc9dd0effd784ba3d31298b14785f8e9a93db into 0f113e662f7b8191c82cff5a819f915724978f2e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-4f57dc63bab607e2c338dd2d4bdf5f2cf642860c)\n\n**new alerts:**\n\n* 3 for Unused import",
"This pull request **introduces 3 alerts** when merging f15c40a34c35db5d8fac397f8e5be2752daabaa0 into 0f113e662f7b8191c82cff5a819f915724978f2e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-85e77b066a531fa209e39a3d93b7c1b203af5f65)\n\n**new alerts:**\n\n* 3 for Unused import"
] | 2020-11-05T01:15:39 | 2021-11-05T19:32:42 | 2020-12-17T07:20:23 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1415",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1415",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1415.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1415.patch",
"merged_at": null
} | PR adding a classical (from Attention-is-all-you-need paper) NMT model | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1415/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1415/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1414 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1414/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1414/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1414/events | https://github.com/NVIDIA/NeMo/pull/1414 | 736,416,147 | MDExOlB1bGxSZXF1ZXN0NTE1NjMzODgy | 1,414 | Update checkpointing to support model parallel | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging f0345ed4dbf94798b1375f4798b997ebd1115f22 into 99ed9c2ae5c840a8723f236530ccb1abe8c75377 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-fa795c7f3253cb10f0b090983cf620dbc7125e51)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **introduces 1 alert** when merging d7da994dbae189c1e02a6f51efe924046ca2f03b into 99ed9c2ae5c840a8723f236530ccb1abe8c75377 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-9bea83783937cb7e953f7b8a603b137f54e21238)\n\n**new alerts:**\n\n* 1 for Unused import",
"This pull request **fixes 2 alerts** when merging 0e0f7cfcf8579705f3e7dad5d6f54340e0ecb8f0 into 4de7bd8394fb7e809e04ee8bcadcf9a1711ef497 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-8aa805f1579286cee43cd4b310d05532f97d9e0b)\n\n**fixed alerts:**\n\n* 2 for Unused local variable"
] | 2020-11-04T20:52:45 | 2021-03-02T22:12:07 | 2021-01-18T19:52:24 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1414",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1414",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1414.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1414.patch",
"merged_at": null
} | TODO | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1414/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1414/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1413 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1413/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1413/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1413/events | https://github.com/NVIDIA/NeMo/pull/1413 | 736,306,158 | MDExOlB1bGxSZXF1ZXN0NTE1NTQzMzA1 | 1,413 | bugfixes from v1.0.0b2 | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-04T17:44:23 | 2020-11-04T18:13:40 | 2020-11-04T18:13:40 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1413",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1413",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1413.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1413.patch",
"merged_at": null
} | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1413/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1413/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1412 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1412/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1412/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1412/events | https://github.com/NVIDIA/NeMo/issues/1412 | 736,257,913 | MDU6SXNzdWU3MzYyNTc5MTM= | 1,412 | [Question] Does TTS modules (for example Tacotron2) support fine-tuning like ASR ? | {
"login": "khalilRhouma",
"id": 35801846,
"node_id": "MDQ6VXNlcjM1ODAxODQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/35801846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/khalilRhouma",
"html_url": "https://github.com/khalilRhouma",
"followers_url": "https://api.github.com/users/khalilRhouma/followers",
"following_url": "https://api.github.com/users/khalilRhouma/following{/other_user}",
"gists_url": "https://api.github.com/users/khalilRhouma/gists{/gist_id}",
"starred_url": "https://api.github.com/users/khalilRhouma/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/khalilRhouma/subscriptions",
"organizations_url": "https://api.github.com/users/khalilRhouma/orgs",
"repos_url": "https://api.github.com/users/khalilRhouma/repos",
"events_url": "https://api.github.com/users/khalilRhouma/events{/privacy}",
"received_events_url": "https://api.github.com/users/khalilRhouma/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Hello, any update on this issue?\r\n\r\nI have the same problem, I would like to perform a fine-tuning of Tacotron 2 model with a different set of labels (characters/phonemes in input to the model).\r\n\r\nI'm also interested in simple fine-tuning without changing the set of labels, is this feature already available in Nemo? I didn't find the option to start the training from a checkpoint.\r\n\r\nThanks!",
"Tacotron is not easily finetunable, we suggest using other supported TTS models which support finetuning. "
] | 2020-11-04T16:36:47 | 2022-01-18T18:40:42 | 2022-01-18T18:40:06 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Is fine-tuning with TTS not supported yet?**
Am working on transfer learning with TTS models From English to the Arabic language. I saw an example of fine-tuning using ASR Quartznet model mentioned [here](https://github.com/NVIDIA/NeMo/issues/1107#issuecomment-686005952).
I am using this pre-trained model:
`tacotron2= nemo_tts.models.Tacotron2Model.from_pretrained("Tacotron2-22050Hz")`
I found that Tacotron model doesn't support `model.change_vocabulary()` and it raises receptions when I am using `model.setup_training_data()`.
Nemo version: 1.0.0b1
Method of NeMo install: `pip install nemo_toolkit[all]==1.0.0b1`
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1412/reactions",
"total_count": 5,
"+1": 5,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1412/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1411 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1411/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1411/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1411/events | https://github.com/NVIDIA/NeMo/issues/1411 | 735,978,617 | MDU6SXNzdWU3MzU5Nzg2MTc= | 1,411 | When i use GPU to infer, everything is ok. But when i choose CPU, it turn out runtime error: code is too big. I have no idea what is it mean. Did anybody know how can i fix it. | {
"login": "tianchaolangzi",
"id": 29921950,
"node_id": "MDQ6VXNlcjI5OTIxOTUw",
"avatar_url": "https://avatars.githubusercontent.com/u/29921950?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tianchaolangzi",
"html_url": "https://github.com/tianchaolangzi",
"followers_url": "https://api.github.com/users/tianchaolangzi/followers",
"following_url": "https://api.github.com/users/tianchaolangzi/following{/other_user}",
"gists_url": "https://api.github.com/users/tianchaolangzi/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tianchaolangzi/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tianchaolangzi/subscriptions",
"organizations_url": "https://api.github.com/users/tianchaolangzi/orgs",
"repos_url": "https://api.github.com/users/tianchaolangzi/repos",
"events_url": "https://api.github.com/users/tianchaolangzi/events{/privacy}",
"received_events_url": "https://api.github.com/users/tianchaolangzi/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Please upgrade to 1.0.0b* version of NeMo. See https://github.com/NVIDIA/NeMo/issues/1091 for a discussion on the latest changes."
] | 2020-11-04T10:02:02 | 2020-12-04T02:39:11 | 2020-12-04T02:38:33 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I set batch_size = 1, it still shows runtime error.
But when my test_set only have one wav, it can run successfully.
part of my code:
neural_factory = nemo.core.NeuralModuleFactory(
backend=nemo.core.Backend.PyTorch,
#local_rank=args.local_rank,
#optimization_level=args.amp_opt_level,
placement=nemo.core.DeviceType.CPU,
)
evaluated_tensors = neural_factory.infer(tensors=eval_tensors, checkpoint_dir=load_dir)
The error message:
/opt/conda/lib/python3.6/site-packages/torchvision/io/_video_opt.py:17: UserWarning: video reader based on ffmpeg c++ ops not available
warnings.warn("video reader based on ffmpeg c++ ops not available")
[NeMo I 2020-11-04 09:53:12 collections:152] Dataset loaded with 12203 files totalling 6.17 hours
[NeMo I 2020-11-04 09:53:12 collections:153] 0 files were filtered totalling 0.00 hours
[NeMo I 2020-11-04 09:53:12 stt_infer:138] Evaluating 12203 examples
[NeMo I 2020-11-04 09:53:12 features:144] PADDING: 16
[NeMo I 2020-11-04 09:53:12 features:152] STFT using conv
[NeMo I 2020-11-04 09:53:13 stt_infer:151] ================================
[NeMo I 2020-11-04 09:53:13 stt_infer:152] Number of parameters in encoder: 18894656
[NeMo I 2020-11-04 09:53:13 stt_infer:153] Number of parameters in decoder: 29725
[NeMo I 2020-11-04 09:53:13 stt_infer:154] Total number of parameters in model: 18924381
[NeMo I 2020-11-04 09:53:13 stt_infer:155] ================================
[NeMo I 2020-11-04 09:53:13 actions:1548] Restoring JasperEncoder from asr_checkpoint/JasperEncoder.pt
[NeMo I 2020-11-04 09:53:13 actions:1548] Restoring JasperDecoderForCTC from asr_checkpoint/JasperDecoderForCTC.pt
[NeMo I 2020-11-04 09:53:13 actions:673] Evaluating batch 0 out of 12203
Traceback (most recent call last):
File "stt_infer.py", line 221, in <module>
main()
File "stt_infer.py", line 168, in main
evaluated_tensors = neural_factory.infer(tensors=eval_tensors, checkpoint_dir=load_dir)
File "/NeMo/nemo/core/neural_factory.py", line 421, in infer
modules_to_restore=modules_to_restore,
File "/NeMo/nemo/backends/pytorch/actions.py", line 1575, in infer
offload_to_cpu=offload_to_cpu,
File "/NeMo/nemo/backends/pytorch/actions.py", line 700, in _infer
use_cache=use_cache,
File "/NeMo/nemo/backends/pytorch/actions.py", line 362, in __nm_graph_forward_pass
new_tensors = pmodule(force_pt=True, **call_set)
File "/NeMo/nemo/backends/pytorch/nm.py", line 160, in __call__
return self.forward(*input, **kwargs)
File "/opt/conda/lib/python3.6/site-packages/torch/autograd/grad_mode.py", line 49, in decorate_no_grad
return func(*args, **kwargs)
File "/NeMo/nemo/collections/asr/audio_preprocessing.py", line 93, in forward
processed_signal = self.get_features(input_signal, length)
File "/NeMo/nemo/collections/asr/audio_preprocessing.py", line 385, in get_features
return self.featurizer(input_signal, length)
File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/opt/conda/lib/python3.6/site-packages/torch/autograd/grad_mode.py", line 49, in decorate_no_grad
return func(*args, **kwargs)
File "/NeMo/nemo/collections/asr/parts/features.py", line 252, in forward
x = self.stft(x)
File "/opt/conda/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/NeMo/nemo/collections/asr/parts/features.py", line 160, in forward
return super(STFTPatch, self).transform(input_data)[0]
File "/opt/conda/lib/python3.6/site-packages/torch_stft/stft.py", line 86, in transform
padding=0)
RuntimeError: code is too big | {
"login": "tianchaolangzi",
"id": 29921950,
"node_id": "MDQ6VXNlcjI5OTIxOTUw",
"avatar_url": "https://avatars.githubusercontent.com/u/29921950?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tianchaolangzi",
"html_url": "https://github.com/tianchaolangzi",
"followers_url": "https://api.github.com/users/tianchaolangzi/followers",
"following_url": "https://api.github.com/users/tianchaolangzi/following{/other_user}",
"gists_url": "https://api.github.com/users/tianchaolangzi/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tianchaolangzi/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tianchaolangzi/subscriptions",
"organizations_url": "https://api.github.com/users/tianchaolangzi/orgs",
"repos_url": "https://api.github.com/users/tianchaolangzi/repos",
"events_url": "https://api.github.com/users/tianchaolangzi/events{/privacy}",
"received_events_url": "https://api.github.com/users/tianchaolangzi/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1411/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1411/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1410 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1410/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1410/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1410/events | https://github.com/NVIDIA/NeMo/pull/1410 | 735,974,043 | MDExOlB1bGxSZXF1ZXN0NTE1MjY4MjU5 | 1,410 | Model save/restore based on EFF.NeMoArchive | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-04T09:55:52 | 2020-11-05T04:57:34 | 2020-11-05T03:42:31 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1410",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1410",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1410.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1410.patch",
"merged_at": "2020-11-05T03:42:31"
} | Rebased branch, reopened the PR | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1410/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1410/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1409 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1409/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1409/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1409/events | https://github.com/NVIDIA/NeMo/pull/1409 | 735,911,011 | MDExOlB1bGxSZXF1ZXN0NTE1MjE2NzAy | 1,409 | Patch exp_manager.py for PTL 1.0.5 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-04T08:21:52 | 2020-11-04T20:42:59 | 2020-11-04T17:33:47 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1409",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1409",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1409.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1409.patch",
"merged_at": "2020-11-04T17:33:47"
} | Moved from previous PR due to wrong base branch.
Signed-off-by: smajumdar <[email protected]> | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1409/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1409/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1408 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1408/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1408/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1408/events | https://github.com/NVIDIA/NeMo/issues/1408 | 735,896,038 | MDU6SXNzdWU3MzU4OTYwMzg= | 1,408 | [Question] Conformer pre-trained models | {
"login": "hoangtuanvu",
"id": 29234461,
"node_id": "MDQ6VXNlcjI5MjM0NDYx",
"avatar_url": "https://avatars.githubusercontent.com/u/29234461?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hoangtuanvu",
"html_url": "https://github.com/hoangtuanvu",
"followers_url": "https://api.github.com/users/hoangtuanvu/followers",
"following_url": "https://api.github.com/users/hoangtuanvu/following{/other_user}",
"gists_url": "https://api.github.com/users/hoangtuanvu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/hoangtuanvu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/hoangtuanvu/subscriptions",
"organizations_url": "https://api.github.com/users/hoangtuanvu/orgs",
"repos_url": "https://api.github.com/users/hoangtuanvu/repos",
"events_url": "https://api.github.com/users/hoangtuanvu/events{/privacy}",
"received_events_url": "https://api.github.com/users/hoangtuanvu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"I would also be very interested!",
"Hi, first thanks for developing this repo. It is quite useful. I am also interested in the pretrained conformer model. It would be helpful if this could also be added to ngc\r\n\r\nThanks,",
"We have the plan to release publicly some checkpoints with Conformer-CTC, and maybe even Conformer-RNNT in our next release in March.",
"> We have the plan to release publicly some checkpoints with Conformer-CTC, and maybe even Conformer-RNNT in our next release in March.\r\n\r\nThanks Vahid, this would be very helpful",
"Thanks @VahidooX",
"I see there are now pretrained conformers for the CTCEncDecBPE models [here](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/results.html#automatic-speech-recognition-models). Are there any checkpoints anywhere for the char version? I found the yaml file for training conformer + char but can't seem to find any pretrained models or checkpoints. Thanks!",
"We do not have any plan to release any publicly available checkpoints trained with char-level encoding. In our experiments, accuracy of char-level models was worse than BPE-level models. Generally when you use downsampling, BPE-based models may work better. The current released checkpoints for Conformer-CTC are trained with just LS, and we have the plan to release some better checkpoints trained on larger datasets in the following month.",
"@VahidooX , any plans on releasing Conformer-RNNT pre-trained weights? It would be helpful if you could share any sort of rough timeline as to when can we expect it. ",
"It is on our roadmap and very likely that we release some RNNT models by the end of July, maybe even sooner. We are also going to release some Conformer-CTC models trained on larger datasets this month.",
"Conformer-CTC checkpoints published:\r\nhttps://ngc.nvidia.com/catalog/collections/nvidia:nemo_asr",
"Is there anywhere a public table with librispeech test-other results?\r\n\r\nModel Cards at https://ngc.nvidia.com/catalog/collections/nvidia:nemo_asr require nvidia login. On the side, I think this is baseless restriction of access.",
"You may find the results on test-other in their corresponding model cards. Unfortunately we don't have a table with all results in the same place. You don't need to create accounts to get access to NGC, there is a button labelled \"guest\" on left bottom side of the login page. If you do it just once, you get access to all public pages on NGC."
] | 2020-11-04T07:56:35 | 2021-10-25T17:40:40 | 2021-06-10T21:59:48 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Thanks authors for publishing ASR sources. I know that QuartzNet & Jasper are comparable to some SOTA model for S2T tasks. Recently, Google published "Conformer" encoder network and achieved very good performance on LibriSpeech. I saw that, that network source is released on new Nemo version. Could you please release pre-trained model for this? | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1408/reactions",
"total_count": 14,
"+1": 11,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 3,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1408/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1407 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1407/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1407/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1407/events | https://github.com/NVIDIA/NeMo/pull/1407 | 735,808,002 | MDExOlB1bGxSZXF1ZXN0NTE1MTMwNTU4 | 1,407 | Patch exp_manager.py for PTL 1.0.5 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-04T04:47:50 | 2020-11-04T08:25:16 | 2020-11-04T08:16:45 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1407",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1407",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1407.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1407.patch",
"merged_at": null
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1407/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1407/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1406 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1406/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1406/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1406/events | https://github.com/NVIDIA/NeMo/pull/1406 | 735,736,545 | MDExOlB1bGxSZXF1ZXN0NTE1MDc0MDEx | 1,406 | Fix another hanging bug in text classifiction model. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging 260e83eac7de1938e07b5665c52b5d5f26906d50 into 536cf58e010a5ce389ed16ce52acce80e5524a1d - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-7330a0019def9291ed2c5484a46b6fc4616b32f2)\n\n**new alerts:**\n\n* 2 for Unused import"
] | 2020-11-04T01:07:20 | 2020-11-05T01:48:34 | 2020-11-05T01:48:34 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1406",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1406",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1406.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1406.patch",
"merged_at": "2020-11-05T01:48:34"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1406/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1406/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1405 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1405/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1405/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1405/events | https://github.com/NVIDIA/NeMo/pull/1405 | 735,653,536 | MDExOlB1bGxSZXF1ZXN0NTE1MDA2NTY3 | 1,405 | Fix smoothed cross entropy loss | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T21:41:45 | 2020-11-04T15:08:18 | 2020-11-04T15:08:11 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1405",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1405",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1405.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1405.patch",
"merged_at": "2020-11-04T15:08:11"
} | `SmoothedCrossEntropyLoss` is expecting log probs but was using logits type and logits name for variables.
Changed logits to log_probs and updated types. | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1405/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1405/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1404 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1404/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1404/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1404/events | https://github.com/NVIDIA/NeMo/pull/1404 | 735,533,820 | MDExOlB1bGxSZXF1ZXN0NTE0OTA5MzU3 | 1,404 | Text2Sparql v1.0.0b2 bug fixes. | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T18:09:47 | 2020-11-03T18:46:33 | 2020-11-03T18:46:25 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1404",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1404",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1404.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1404.patch",
"merged_at": "2020-11-03T18:46:25"
} | Pinned notebook to v1.0.0b2.
Fixed a small typo.
| {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1404/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1404/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1403 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1403/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1403/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1403/events | https://github.com/NVIDIA/NeMo/pull/1403 | 735,043,706 | MDExOlB1bGxSZXF1ZXN0NTE0NTAzOTcz | 1,403 | QAModel exportable | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T06:02:48 | 2020-11-03T06:21:39 | 2020-11-03T06:21:39 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1403",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1403",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1403.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1403.patch",
"merged_at": "2020-11-03T06:21:39"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1403/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1403/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1402 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1402/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1402/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1402/events | https://github.com/NVIDIA/NeMo/pull/1402 | 734,980,472 | MDExOlB1bGxSZXF1ZXN0NTE0NDUwOTY2 | 1,402 | Fixed hanging bug of text classifcation. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T03:08:48 | 2020-11-03T03:53:31 | 2020-11-03T03:53:31 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1402",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1402",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1402.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1402.patch",
"merged_at": "2020-11-03T03:53:31"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1402/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1402/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1401 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1401/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1401/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1401/events | https://github.com/NVIDIA/NeMo/pull/1401 | 734,956,227 | MDExOlB1bGxSZXF1ZXN0NTE0NDMxNzQ2 | 1,401 | Fix to support any Python version | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"please run ```python setup.py style --fix``` to fix style and let CI run"
] | 2020-11-03T01:49:18 | 2020-11-04T02:06:47 | 2020-11-04T02:06:47 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1401",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1401",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1401.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1401.patch",
"merged_at": "2020-11-04T02:06:47"
} | Signed-off-by: Sergei Nikolaev <[email protected]>
Fix to support any Python version | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1401/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1401/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1400 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1400/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1400/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1400/events | https://github.com/NVIDIA/NeMo/pull/1400 | 734,953,905 | MDExOlB1bGxSZXF1ZXN0NTE0NDI5ODUz | 1,400 | Fixed the evaluation trainer in text classification model. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T01:41:25 | 2020-11-03T02:53:07 | 2020-11-03T02:53:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1400",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1400",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1400.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1400.patch",
"merged_at": "2020-11-03T02:53:07"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1400/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1400/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1399 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1399/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1399/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1399/events | https://github.com/NVIDIA/NeMo/pull/1399 | 734,942,304 | MDExOlB1bGxSZXF1ZXN0NTE0NDIwNTU1 | 1,399 | Add check val frequency to BPE configs | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-03T01:04:02 | 2020-11-03T19:34:34 | 2020-11-03T06:10:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1399",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1399",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1399.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1399.patch",
"merged_at": "2020-11-03T06:10:59"
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1399/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1399/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1398 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1398/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1398/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1398/events | https://github.com/NVIDIA/NeMo/pull/1398 | 734,909,993 | MDExOlB1bGxSZXF1ZXN0NTE0MzkzOTY2 | 1,398 | Remove call to optimizer from the model initialization and added upda… | {
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@vladgets do you still want to merge this?",
"> @vladgets do you still want to merge this?\r\n\r\nYes, I do",
"> > @vladgets do you still want to merge this?\r\n> \r\n> Yes, I do\r\n\r\nShould this go to main or v1.0.0b2 @okuchaiev? It's needed for tlt"
] | 2020-11-02T23:35:24 | 2020-11-04T19:08:14 | 2020-11-04T19:07:59 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1398",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1398",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1398.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1398.patch",
"merged_at": "2020-11-04T19:07:59"
} | Removed call to optimizer from the model initialization and added update data directory function | {
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1398/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1398/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1397 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1397/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1397/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1397/events | https://github.com/NVIDIA/NeMo/pull/1397 | 734,810,759 | MDExOlB1bGxSZXF1ZXN0NTE0MzA5ODkz | 1,397 | NeMo Model EFF Strategies | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-11-02T20:51:41 | 2020-11-04T09:54:12 | 2020-11-04T09:53:53 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1397",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1397",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1397.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1397.patch",
"merged_at": null
} | Signed-off-by: Tomasz Kornuta <[email protected]> | {
"login": "tkornuta-nvidia",
"id": 56979727,
"node_id": "MDQ6VXNlcjU2OTc5NzI3",
"avatar_url": "https://avatars.githubusercontent.com/u/56979727?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tkornuta-nvidia",
"html_url": "https://github.com/tkornuta-nvidia",
"followers_url": "https://api.github.com/users/tkornuta-nvidia/followers",
"following_url": "https://api.github.com/users/tkornuta-nvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/tkornuta-nvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tkornuta-nvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tkornuta-nvidia/subscriptions",
"organizations_url": "https://api.github.com/users/tkornuta-nvidia/orgs",
"repos_url": "https://api.github.com/users/tkornuta-nvidia/repos",
"events_url": "https://api.github.com/users/tkornuta-nvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/tkornuta-nvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1397/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1397/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1396 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1396/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1396/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1396/events | https://github.com/NVIDIA/NeMo/pull/1396 | 734,788,611 | MDExOlB1bGxSZXF1ZXN0NTE0MjkyMDQy | 1,396 | Text_Classification_Sentiment_Analysis.ipynb fix | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"code format check fails - see CI output"
] | 2020-11-02T20:14:13 | 2020-11-05T01:49:41 | 2020-11-05T01:49:41 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1396",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1396",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1396.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1396.patch",
"merged_at": "2020-11-05T01:49:40"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1396/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1396/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1395 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1395/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1395/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1395/events | https://github.com/NVIDIA/NeMo/issues/1395 | 734,311,552 | MDU6SXNzdWU3MzQzMTE1NTI= | 1,395 | [Question] How to train a ASR model finetuned with RIR and noise augmentation, just like `QuartzNet15x5NR-En`? | {
"login": "Honghe",
"id": 1092722,
"node_id": "MDQ6VXNlcjEwOTI3MjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1092722?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Honghe",
"html_url": "https://github.com/Honghe",
"followers_url": "https://api.github.com/users/Honghe/followers",
"following_url": "https://api.github.com/users/Honghe/following{/other_user}",
"gists_url": "https://api.github.com/users/Honghe/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Honghe/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Honghe/subscriptions",
"organizations_url": "https://api.github.com/users/Honghe/orgs",
"repos_url": "https://api.github.com/users/Honghe/repos",
"events_url": "https://api.github.com/users/Honghe/events{/privacy}",
"received_events_url": "https://api.github.com/users/Honghe/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"Here are some scripts that can help you create datasets for RIR and noise augmentation:\r\nhttps://github.com/NVIDIA/NeMo/blob/main/scripts/get_openslr_rir.py (downloads and creates a json manifest for RIRs from openslr)\r\n\r\nSome scripts to download noise samples from freesound are here: https://github.com/NVIDIA/NeMo/tree/main/scripts/freesound_download_resample\r\n\r\nFor training the model, we found fine-tuning a model trained on \"clean\" data to be quite effective. Here is an example config file that can get you started with training a noise robust model: https://github.com/NVIDIA/NeMo/blob/main/examples/asr/experimental/configs/quartznet_15x5_aug.yaml . Look for \"augmentor\" settings in this config file. \r\n\r\n",
"@okuchaiev @jbalam-nv Thanks! And I have made a pull request of RIR perturbation example in the tutorial. https://github.com/NVIDIA/NeMo/pull/1429",
"Hello,\r\nim tryin to implement RIR aug, just like suggested in the [tutorial](https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/05_Online_Noise_Augmentation.ipynb\r\n)\r\nBut when i run the cell:\r\n```\r\nrir_data_path = '.'\r\n!python ../../scripts/get_openslr_rir.py --data_root {rir_data_path}\r\nrir_manifest_path = os.path.join(rir_data_path, 'processed', 'rir.json')\r\n!head -n 3 {rir_manifest_path}\r\n```\r\nIt returns:\r\n\r\n```\r\npython3: can't open file '../../scripts/get_openslr_rir.py': [Errno 2] No such file or directory\r\nhead: cannot open './processed/rir.json' for reading: No such file or directory\r\n\r\n```\r\nSo, i downloaded get_openslr_rir.py script manually from [here](https://raw.githubusercontent.com/NVIDIA/NeMo/main/scripts/get_openslr_rir.py)\r\n\r\nAnd the cell to run became:\r\n```\r\nrir_data_path = '.'\r\n!python get_openslr_rir.py --data_root {rir_data_path}\r\nrir_manifest_path = os.path.join(rir_data_path, 'processed', 'rir.json')\r\n!head -n 3 {rir_manifest_path}\r\n```\r\n\r\nBut still get an error after downloading and extracting the file:\r\n\r\n```\r\nINFO:root:\r\n\r\nWorking on: slr28\r\nINFO:root:Getting slr28\r\nINFO:root:./slr28.zip does not exist. Downloading ...\r\nINFO:root:Downloaded ./slr28.zip.\r\nINFO:root:Extracting slr28\r\nINFO:root:Processing slr28\r\nTraceback (most recent call last):\r\n File \"get_openslr_rir.py\", line 139, in <module>\r\n main()\r\n File \"get_openslr_rir.py\", line 133, in main\r\n os.path.join(os.path.join(data_root, \"processed\", \"rir.json\")),\r\n File \"get_openslr_rir.py\", line 105, in __process_data\r\n with open(manifest_file, 'w') as man_f:\r\nFileNotFoundError: [Errno 2] No such file or directory: './processed/rir.json'\r\nhead: cannot open './processed/rir.json' for reading: No such file or directory\r\n\r\n```\r\n\r\nIs this a known issue? Or i'm just bad af and i don't know what to do? I can't even look at how rir manifest are structured. And this is realy important for me to add RiRAug in my ASR pipeline.\r\n\r\nAppreciate some help\r\nThanks in advance\r\n",
"@lodm94 Do you have permission to create this folder and file? I was able to download and convert the data without any problems.\r\n"
] | 2020-11-02T08:44:46 | 2021-06-19T04:47:18 | 2021-06-19T04:47:18 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe your question**
How to train a ASR model finetuned with RIR and noise augmentation, just like `QuartzNet15x5NR-En`?
I find that there is a pretrained `QuartzNet15x5NR-En` model on NGC, and a config file for augmentation `/home/jack/a/PycharmProjects/NeMo/examples/asr/experimental/configs/quartznet_15x5_aug.yaml`.
```
nemo_asr.models.EncDecCTCModel.list_available_models()
Out[2]:
PretrainedModelInfo(pretrained_model_name='QuartzNet15x5NR-En', description='QuartzNet15x5Base-En was finetuned with RIR and noise augmentation to make it more robust to noise. This model should be preferred for noisy speech transcription. This model achieves a WER of 3.96% on LibriSpeech dev-clean and a WER of 10.14% on dev-other.', location='https://api.ngc.nvidia.com/v2/models/nvidia/nemospeechmodels/versions/1.0.0a5/files/QuartzNet15x5NR-En.nemo', class_=None),
```
But I can't find the recipe for training the RIR and noise augmented model `QuartzNet15x5NR-En`.
So it will be meaningful if NeMo can provide the the recipe.
**Environment overview (please complete the following information)**
- Environment location: Bare-metal
- Method of NeMo install: from source main(75774cf)
**Environment details**
- OS version: Ubuntu 20.04
- PyTorch version: 1.6
- Python version: 3.8 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1395/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1395/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1394 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1394/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1394/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1394/events | https://github.com/NVIDIA/NeMo/issues/1394 | 734,213,887 | MDU6SXNzdWU3MzQyMTM4ODc= | 1,394 | How to use/port quartzNet model on mobile? [Question] | {
"login": "rahul-hiker",
"id": 73813396,
"node_id": "MDQ6VXNlcjczODEzMzk2",
"avatar_url": "https://avatars.githubusercontent.com/u/73813396?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rahul-hiker",
"html_url": "https://github.com/rahul-hiker",
"followers_url": "https://api.github.com/users/rahul-hiker/followers",
"following_url": "https://api.github.com/users/rahul-hiker/following{/other_user}",
"gists_url": "https://api.github.com/users/rahul-hiker/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rahul-hiker/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rahul-hiker/subscriptions",
"organizations_url": "https://api.github.com/users/rahul-hiker/orgs",
"repos_url": "https://api.github.com/users/rahul-hiker/repos",
"events_url": "https://api.github.com/users/rahul-hiker/events{/privacy}",
"received_events_url": "https://api.github.com/users/rahul-hiker/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | true | null | [] | null | [
"> Can I deploy NeMo trained ASR on mobile platform (Android/iOS)? \r\n- Yes!\r\n\r\n> If yes then what are the steps?\r\n- Every NeMo model is inherited from ```torch.nn.Module``` , therefore you should checkout Pytorch tutorials about deploying models on mobile. For example, you can export QuartzNet ASR models into .onnx file for deployment.\r\n\r\n> Is there a working demo?\r\n- Unfortunately, we don't have an open-sourced working demo yet.",
"@okuchaiev Thanks for the reply.\r\n\r\nI am looking for some answers in order to deploy the NeMo model on mobile.\r\n\r\n1. How to export Nemo quartzNet model to .onnx?\r\n2. When we deploy it on mobile then what are the libraries required (does it have some dependency on 3rd part libraries apart from pytorch code)?\r\n3. What will be inference code in mobile?",
"@rahul-hiker you can follow the instruction: https://pytorch.org/mobile/home/\r\nI hope it helps you deploying Nemo model on mobile"
] | 2020-11-02T06:08:42 | 2021-02-10T23:29:38 | 2021-02-10T23:29:38 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Can I deploy NeMo trained ASR on mobile platform (Android/iOS)? If yes then what are the steps? Is there a working demo? | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1394/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1394/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1393 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1393/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1393/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1393/events | https://github.com/NVIDIA/NeMo/issues/1393 | 734,203,099 | MDU6SXNzdWU3MzQyMDMwOTk= | 1,393 | [Question] How do I infer the speaker recognition model/module at 1.0.0b1. | {
"login": "uditdixit11",
"id": 45828611,
"node_id": "MDQ6VXNlcjQ1ODI4NjEx",
"avatar_url": "https://avatars.githubusercontent.com/u/45828611?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/uditdixit11",
"html_url": "https://github.com/uditdixit11",
"followers_url": "https://api.github.com/users/uditdixit11/followers",
"following_url": "https://api.github.com/users/uditdixit11/following{/other_user}",
"gists_url": "https://api.github.com/users/uditdixit11/gists{/gist_id}",
"starred_url": "https://api.github.com/users/uditdixit11/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/uditdixit11/subscriptions",
"organizations_url": "https://api.github.com/users/uditdixit11/orgs",
"repos_url": "https://api.github.com/users/uditdixit11/repos",
"events_url": "https://api.github.com/users/uditdixit11/events{/privacy}",
"received_events_url": "https://api.github.com/users/uditdixit11/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"@uditdixit11, \r\nIf you have finetuned or trained all together with a new speaker recognition model, you could test or infer by two steps as shown similarly in [this script](https://github.com/NVIDIA/NeMo/blob/main/examples/asr/speech_to_text_infer.py) . \r\n1. set up test data and \r\n2. get log_prob by calling forward loop of trained speaker model. \r\nIf it's confusing let me know, I will add a infer script to current examples. ",
"Thanks @nithinraok\r\n\r\nPlease add the infer script to the current examples. It will be really helpful to us.",
"@nithinraok Would appreciate an infer script as well, working on it myself right now.",
"Did you try passing test_ds to yaml file to speaker_reco.py by commenting out trainer.fit with your trained .nemo model? ",
"Anyway will post a script very soon for your reference. ",
"Somehow missed your response @uditdixit11, please find the script requested at this [link](https://github.com/NVIDIA/NeMo/blob/afbe52cb726eeb0c8262779299f4cc4354a9625d/examples/speaker_recognition/speaker_reco_infer.py). Let me how it goes. @ShantanuNair ",
"Hi! It appears in the script setting up test data loader and load from test data is a necessary step, in which process audio data were somehow converted to a tensor and can be passed to the `forward()` function as an argument, I was wondering if there is a way to get around this? For example, I just have a running program that takes input audio streams, might there be a way to process the audio stream to a required tensor form and pass it directly to the `forward()` function without having to write to an intermediate test manifest and load from the test data? Has anyone tried this? Help would be appreciated! ",
"You may look at this discussion to get an idea: https://github.com/NVIDIA/NeMo/discussions/2248 ",
"> You may look at this discussion to get an idea: #2248\r\n\r\nThank you! @nithinraok "
] | 2020-11-02T05:48:02 | 2021-06-08T08:15:28 | 2021-01-25T07:29:23 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I have trained a speaker recognition model and save the checkpoints. Even can restore it using the LightningModule.load_from_checkpoint() class method. But couldn't able to perform the next steps after it.
Please suggest how do I perform the next steps i.e then predict the speaker on unseen data using trained model at at 1.0.0b1 version ?
| {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1393/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1393/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1392 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1392/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1392/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1392/events | https://github.com/NVIDIA/NeMo/pull/1392 | 733,956,057 | MDExOlB1bGxSZXF1ZXN0NTEzNjE0Mjg3 | 1,392 | Information Retrieval Tutorial | {
"login": "AlexGrinch",
"id": 8689095,
"node_id": "MDQ6VXNlcjg2ODkwOTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/8689095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AlexGrinch",
"html_url": "https://github.com/AlexGrinch",
"followers_url": "https://api.github.com/users/AlexGrinch/followers",
"following_url": "https://api.github.com/users/AlexGrinch/following{/other_user}",
"gists_url": "https://api.github.com/users/AlexGrinch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AlexGrinch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AlexGrinch/subscriptions",
"organizations_url": "https://api.github.com/users/AlexGrinch/orgs",
"repos_url": "https://api.github.com/users/AlexGrinch/repos",
"events_url": "https://api.github.com/users/AlexGrinch/events{/privacy}",
"received_events_url": "https://api.github.com/users/AlexGrinch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@AlexGrinch can you please fix the PR so that CI passes ",
"@AlexGrinch code style",
"please fix code formatting to unblock CI:\r\n```\r\npython setup.py style --fix\r\n```"
] | 2020-11-01T12:35:52 | 2021-01-13T21:48:44 | 2021-01-13T21:48:43 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1392",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1392",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1392.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1392.patch",
"merged_at": "2021-01-13T21:48:43"
} | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1392/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1392/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1391 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1391/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1391/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1391/events | https://github.com/NVIDIA/NeMo/issues/1391 | 733,925,734 | MDU6SXNzdWU3MzM5MjU3MzQ= | 1,391 | about japanese | {
"login": "guijuzhejiang",
"id": 57936895,
"node_id": "MDQ6VXNlcjU3OTM2ODk1",
"avatar_url": "https://avatars.githubusercontent.com/u/57936895?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/guijuzhejiang",
"html_url": "https://github.com/guijuzhejiang",
"followers_url": "https://api.github.com/users/guijuzhejiang/followers",
"following_url": "https://api.github.com/users/guijuzhejiang/following{/other_user}",
"gists_url": "https://api.github.com/users/guijuzhejiang/gists{/gist_id}",
"starred_url": "https://api.github.com/users/guijuzhejiang/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/guijuzhejiang/subscriptions",
"organizations_url": "https://api.github.com/users/guijuzhejiang/orgs",
"repos_url": "https://api.github.com/users/guijuzhejiang/repos",
"events_url": "https://api.github.com/users/guijuzhejiang/events{/privacy}",
"received_events_url": "https://api.github.com/users/guijuzhejiang/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Yes, NeMo is not English-only. You should be able to train good Japanese models if you have Japanese training data. Of course, knowledge of Japanese helps too :) ",
"> Yes, NeMo is not English-only. You should be able to train good Japanese models if you have Japanese training data. Of course, knowledge of Japanese helps too :)\r\n\r\nThank you for your answer. It seems that there is no automatic speech recognition pre-training model for Japanese"
] | 2020-11-01T09:56:01 | 2020-11-03T09:07:55 | 2020-11-03T09:07:55 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Thank you very much for opening such a good tool. Is nemo effective for Japanese speech recognition? | {
"login": "guijuzhejiang",
"id": 57936895,
"node_id": "MDQ6VXNlcjU3OTM2ODk1",
"avatar_url": "https://avatars.githubusercontent.com/u/57936895?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/guijuzhejiang",
"html_url": "https://github.com/guijuzhejiang",
"followers_url": "https://api.github.com/users/guijuzhejiang/followers",
"following_url": "https://api.github.com/users/guijuzhejiang/following{/other_user}",
"gists_url": "https://api.github.com/users/guijuzhejiang/gists{/gist_id}",
"starred_url": "https://api.github.com/users/guijuzhejiang/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/guijuzhejiang/subscriptions",
"organizations_url": "https://api.github.com/users/guijuzhejiang/orgs",
"repos_url": "https://api.github.com/users/guijuzhejiang/repos",
"events_url": "https://api.github.com/users/guijuzhejiang/events{/privacy}",
"received_events_url": "https://api.github.com/users/guijuzhejiang/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1391/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1391/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1390 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1390/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1390/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1390/events | https://github.com/NVIDIA/NeMo/pull/1390 | 733,648,008 | MDExOlB1bGxSZXF1ZXN0NTEzMzgyNDQz | 1,390 | [WIP] ASR inference script to output hypotheses in sclite formats | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging 9c213ff48acc513dfd3b02654fa92fdd54b93341 into 75774cf24755bddb393e3b1a4cbcc1921daf8898 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-29dea0dd8c242cb81d2215a8c6323dfb9f2b3347)\n\n**new alerts:**\n\n* 2 for Unused local variable"
] | 2020-10-31T06:26:39 | 2020-11-10T08:01:54 | 2020-11-10T08:01:54 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1390",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1390",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1390.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1390.patch",
"merged_at": null
} | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1390/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1390/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/1389 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1389/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1389/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1389/events | https://github.com/NVIDIA/NeMo/pull/1389 | 733,639,416 | MDExOlB1bGxSZXF1ZXN0NTEzMzc2NjY5 | 1,389 | Safeguard g2p_py import (again) | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-31T05:25:42 | 2020-10-31T19:37:17 | 2020-10-31T05:27:04 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1389",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1389",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1389.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1389.patch",
"merged_at": "2020-10-31T05:27:04"
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1389/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1389/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1388 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1388/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1388/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1388/events | https://github.com/NVIDIA/NeMo/pull/1388 | 733,636,242 | MDExOlB1bGxSZXF1ZXN0NTEzMzc0NDEy | 1,388 | Safeguard g2p_py import | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-31T05:04:32 | 2020-10-31T05:16:36 | 2020-10-31T05:05:33 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1388",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1388",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1388.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1388.patch",
"merged_at": "2020-10-31T05:05:33"
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1388/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1388/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1387 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1387/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1387/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1387/events | https://github.com/NVIDIA/NeMo/pull/1387 | 733,635,009 | MDExOlB1bGxSZXF1ZXN0NTEzMzczNTc3 | 1,387 | Safeguard g2p_py import | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-31T04:56:26 | 2020-10-31T05:05:29 | 2020-10-31T05:01:30 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1387",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1387",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1387.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1387.patch",
"merged_at": null
} | Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1387/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1387/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1386 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1386/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1386/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1386/events | https://github.com/NVIDIA/NeMo/pull/1386 | 733,575,346 | MDExOlB1bGxSZXF1ZXN0NTEzMzMxNjcz | 1,386 | Vad notebook fix v1.0.0b2 | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-30T23:38:10 | 2021-03-11T17:29:10 | 2020-10-30T23:39:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1386",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1386",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1386.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1386.patch",
"merged_at": null
} | bug fix for 3164246 | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1386/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1386/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1385 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1385/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1385/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1385/events | https://github.com/NVIDIA/NeMo/pull/1385 | 733,573,872 | MDExOlB1bGxSZXF1ZXN0NTEzMzMwNDgz | 1,385 | Fix to support any Python version | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-30T23:32:21 | 2020-10-31T00:03:39 | 2020-10-31T00:03:39 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1385",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1385",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1385.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1385.patch",
"merged_at": "2020-10-31T00:03:39"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1385/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1385/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1384 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1384/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1384/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1384/events | https://github.com/NVIDIA/NeMo/pull/1384 | 733,537,747 | MDExOlB1bGxSZXF1ZXN0NTEzMzAyMDk4 | 1,384 | VAD inference update | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 5 alerts** when merging c0ea1dedb51f1613303a24005612c44e29f0b303 into aa4541b91341e0dda5300f1ded5f60cbbc74429a - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-6f4f2f933f3cfd83e9c80db0d46a48fe806dd015)\n\n**new alerts:**\n\n* 3 for Unused import\n* 1 for Variable defined multiple times\n* 1 for Nested loops with same variable",
"This pull request **introduces 5 alerts** when merging d95efb621ad7ec3fff48b316117116e109cd32db into aa4541b91341e0dda5300f1ded5f60cbbc74429a - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-893456f8b20f71b9e8ff2b4f64b83827f7cc5b98)\n\n**new alerts:**\n\n* 3 for Unused import\n* 1 for Variable defined multiple times\n* 1 for Nested loops with same variable",
"This pull request **introduces 5 alerts** when merging 4711223a9807aadbd261fd1b7557d70cf3c0ad42 into 75774cf24755bddb393e3b1a4cbcc1921daf8898 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-ad317e1a2ba8c643811447b8f69be1e135afd3d7)\n\n**new alerts:**\n\n* 3 for Unused import\n* 1 for Variable defined multiple times\n* 1 for Nested loops with same variable",
"please don't final review/merge it.\r\nHas one last thing to update.\r\n\r\nTake care of JOINT of long audio files",
"This pull request **introduces 3 alerts** when merging fcf03b25f2241fc0792b7ffad9426f74a488bf79 into 4de7bd8394fb7e809e04ee8bcadcf9a1711ef497 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-710fb58d744a94eec3e72c926bd7982838af9294)\n\n**new alerts:**\n\n* 2 for Unused local variable\n* 1 for Nested loops with same variable",
"Hey @fayejf, appreciate all the work and progress!! Nemo seems to be, at least to me, one of the most accessible and complete end-to-end ASR systems out there, wrt ASR and downstream NLP tasks. \r\n\r\nSo I've been deep diving into the ASR world and NeMo, and came across this PR when checking to see if Nemo supports diarization. According to https://github.com/NVIDIA/NeMo/issues/1299 , it seems like you don't currently? But from here https://github.com/NVIDIA/NeMo/commit/c7656318c260d5c33422e93af248e314673b36cf - it seems like it might be something you guys are working on already? Would I be correct in thinking `spkr_get_emb.py` has pretty much most of what I need to be able to perform diarization? ",
"> Hey @fayejf, appreciate all the work and progress!! Nemo seems to be, at least to me, one of the most accessible and complete end-to-end ASR systems out there, wrt ASR and downstream NLP tasks.\r\n> \r\n> So I've been deep diving into the ASR world and NeMo, and came across this PR when checking to see if Nemo supports diarization. According to #1299 , it seems like you don't currently? But from here [c765631](https://github.com/NVIDIA/NeMo/commit/c7656318c260d5c33422e93af248e314673b36cf) - it seems like it might be something you guys are working on already? Would I be correct in thinking `spkr_get_emb.py` has pretty much most of what I need to be able to perform diarization?\r\n\r\nHi @ShantanuNair Thanks for your interest in NeMo. We are working on diarization 😄 and currently we have voice activity detection, speaker recognition and speaker verification. For diarization you will also need clustering. \r\nSpeech detection ✅ -> speaker embedding extraction ✅ (yes spkr_get_emb.py) -> clustering 🚗 -> (optional) resegmentation . \r\n \r\nHope this helps.",
"This pull request **introduces 4 alerts** when merging c95ce6b2cd0d98e727550c21e38007101352e9ea into 38c854da3f553b6577d6f8b08490e026b53b7ea3 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-e094b7b495eb1b4f4a5349d14fabe850ed5f1da2)\n\n**new alerts:**\n\n* 2 for Unused local variable\n* 1 for Redundant assignment\n* 1 for Nested loops with same variable",
"@jbalam-nv @nithinraok Please review the changes. Thanks! 🥰 ",
"FYI @ShantanuNair The PR is merged. Feel free to play with the scripts. Questions and feedbacks are very welcome!",
"> FYI @ShantanuNair The PR is merged. Feel free to play with the scripts. Questions and feedbacks are very welcome!\r\n\r\nThank you! Appreciate the openness. Will reach out if needed :)",
"Hey @fayejf i am confused with Speaker diarization. How to perform a simple inference of Speaker diarization on a given wav file and get outputs in formats like:\r\n```\r\nspeaker 1 0.0 to 0.15\r\nspeaker 2 0.15 to 0.30\r\nspeaker 4 0.30 to 0.45\r\nspeaker 1 0.45 to 1.00\r\nspeaker 3 1.00 to 1.15\r\n\r\n```\r\nwhere values on right indicates time of uterance\r\nI am aware of `07_Online_Offline_Microphone_VAD_Demo.ipynb` but there i am getting the result of just the vad. But not in the form of `who spoke when` as mentioned above",
"Hi @jainal09 Thanks for your interest in NeMo.\r\nWe don't have an inference of speaker diarization and we are working on it and will release it soon. 😄 \r\nCurrently, we have voice activity detection, speaker recognition, and speaker verification. For diarization, you will also need clustering.\r\n\r\nSpeech detection ✅ (This PR)-> speaker embedding extraction ✅ (spkr_get_emb.py ,etc.) -> clustering 🚗 (WIP) -> (optional) resegmentation",
"Hey @fayejf thanks for replying.\r\nI really appreciate all the work and progress that you guys are putting in nemo.\r\nI will keep around and eagerly waiting for inference of speaker diarization to get released.\r\nReally excited for this!😊",
" 🥰 Thanks! @jainal09 ❤️",
"Hey @fayejf @nithinraok i also have one more question in the events of overlapped speech will this work to recognize both speakers or will it flag it as new speaker?",
"Thanks for your question! @jainal09 It's currently a work-in-progress. Please wait for the release. Feel free to leave questions but we might not have clear answers now.",
"Sure! Thanks @fayejf I will wait till release 🤞.",
"Hi @fayejf any updates on this?\r\nAlso, I had the same questions that @serg06 mentioned here : https://github.com/NVIDIA/NeMo/issues/1299#issuecomment-739462587\r\nAs to performing speaker diarialization as per the method mentioned by you above:\r\n> Speech detection ✅ (This PR)-> speaker embedding extraction ✅ (spkr_get_emb.py ,etc.) -> clustering 🚗 (WIP) -> (optional) resegmentation\r\n",
"Hey @fayejf pls reply any updates diarialization on the script?",
"@jainal09 Wait. It's holiday season as well, so expect things to slow down."
] | 2020-10-30T21:50:58 | 2020-12-29T06:37:51 | 2020-11-11T21:36:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1384",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1384",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1384.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1384.patch",
"merged_at": "2020-11-11T21:36:07"
} | Update VAD inference to make it easy to combine with Speaker Diarization and more user friendly (user would like txt output)
Add/update:
1. Truncate long audio files to separate json line to avoid CUDA memory issue
2. Frame level inference similar to speech_to_text_infer.py because the frame level streaming is kinda different with asr and speech command.
3. A posterior including overlapped prediction and smoothing filter
4. Write to RTTM like file.
AUDIO -> MANIFEST -> VAD OUTPUT TXT -> VAD RTTM-like TABLE
Signed-off-by: fayejf <[email protected]> | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1384/reactions",
"total_count": 1,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 1
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1384/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1383 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1383/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1383/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1383/events | https://github.com/NVIDIA/NeMo/pull/1383 | 733,510,305 | MDExOlB1bGxSZXF1ZXN0NTEzMjc5ODA2 | 1,383 | QA notebook fix v1.0.0b2 | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-30T20:53:19 | 2020-10-30T22:03:54 | 2020-10-30T22:03:50 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1383",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1383",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1383.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1383.patch",
"merged_at": "2020-10-30T22:03:50"
} | Update `get_huggingface_lm_ model` so that it attempts to use config dict or pretrained model name if config file is specified but doesn't exist.
Update notebook to use fstrings for branch. | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1383/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1383/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1382 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1382/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1382/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1382/events | https://github.com/NVIDIA/NeMo/pull/1382 | 733,493,569 | MDExOlB1bGxSZXF1ZXN0NTEzMjY1NTk3 | 1,382 | fixed the hanging bug of text classification model. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging 0a42bc5c87cce6e6bcbe1ccff4772a6caca25a9c into 17fc2541172dd3d781bfa0b5a1fa41576d4bfe21 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-25e2feefaf1a084c11c7c502283923afe3fa5df8)\n\n**new alerts:**\n\n* 1 for Unused import"
] | 2020-10-30T20:25:45 | 2020-11-02T21:34:43 | 2020-10-30T21:44:44 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1382",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1382",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1382.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1382.patch",
"merged_at": "2020-10-30T21:44:44"
} | Dropped the evaluation with another trainer in the same script. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1382/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1382/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1381 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1381/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1381/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1381/events | https://github.com/NVIDIA/NeMo/issues/1381 | 733,435,003 | MDU6SXNzdWU3MzM0MzUwMDM= | 1,381 | Is there a tutorial for data augmentation using transfer learning on pretrained QuartzNet? | {
"login": "catskillsresearch",
"id": 38082822,
"node_id": "MDQ6VXNlcjM4MDgyODIy",
"avatar_url": "https://avatars.githubusercontent.com/u/38082822?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/catskillsresearch",
"html_url": "https://github.com/catskillsresearch",
"followers_url": "https://api.github.com/users/catskillsresearch/followers",
"following_url": "https://api.github.com/users/catskillsresearch/following{/other_user}",
"gists_url": "https://api.github.com/users/catskillsresearch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/catskillsresearch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/catskillsresearch/subscriptions",
"organizations_url": "https://api.github.com/users/catskillsresearch/orgs",
"repos_url": "https://api.github.com/users/catskillsresearch/repos",
"events_url": "https://api.github.com/users/catskillsresearch/events{/privacy}",
"received_events_url": "https://api.github.com/users/catskillsresearch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Here is a tutorial on data augmentation https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/05_Online_Noise_Augmentation.ipynb "
] | 2020-10-30T18:51:18 | 2021-05-28T08:31:22 | 2021-05-28T08:31:22 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | This doc: https://developer.nvidia.com/blog/develop-smaller-speech-recognition-models-with-nvidias-nemo-framework/ mentions that
> The QuartzNet-15×5 model was trained on the LibriSpeech training dataset with two types of data augmentation: speed perturbation and Cutout
If I transfer-learn the English QuartzNet on another language, is there an easy way to add
- Speed perturbation
- Cutout
- Random background noise
to the training loop? I know how to do this in MONAI but it's not as clear in the NeMo design whether there are audio transforms that can be added into the training loop, that is, I don't see the equivalent in NeMo of the MONAI Compose and transform loader stack: https://docs.monai.io/en/latest/transforms.html | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1381/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1381/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1380 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1380/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1380/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1380/events | https://github.com/NVIDIA/NeMo/issues/1380 | 733,414,734 | MDU6SXNzdWU3MzM0MTQ3MzQ= | 1,380 | Is there a Jupyter tutorial notebook for BERT ASR postprocessing? | {
"login": "catskillsresearch",
"id": 38082822,
"node_id": "MDQ6VXNlcjM4MDgyODIy",
"avatar_url": "https://avatars.githubusercontent.com/u/38082822?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/catskillsresearch",
"html_url": "https://github.com/catskillsresearch",
"followers_url": "https://api.github.com/users/catskillsresearch/followers",
"following_url": "https://api.github.com/users/catskillsresearch/following{/other_user}",
"gists_url": "https://api.github.com/users/catskillsresearch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/catskillsresearch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/catskillsresearch/subscriptions",
"organizations_url": "https://api.github.com/users/catskillsresearch/orgs",
"repos_url": "https://api.github.com/users/catskillsresearch/repos",
"events_url": "https://api.github.com/users/catskillsresearch/events{/privacy}",
"received_events_url": "https://api.github.com/users/catskillsresearch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Actually never mind, I can just train using other text-to-text notebooks as a new language pair, the pair being ASR output to ground truth.",
"Hey, they've explained in other issues - https://github.com/NVIDIA/NeMo/issues/1156, https://github.com/NVIDIA/NeMo/issues/1126 that it hasn't been ported yet, but looks like they have some workarounds to adding a Language Model for incorporating context. They might want to consider adding it to their docs since it seems to pop up often.\r\n\r\n\r\n\r\n> Actually never mind, I can just train using other text-to-text notebooks as a new language pair, the pair being ASR output to ground truth.\r\n\r\nAlso, would you mind elaborating a bit on your process? Did it help with results in your case?",
"Hi @ShantanuNair, they had this script in older releases, they just deleted it without explanation.\r\n\r\nI will elaborate, with this caveat: It didn't help.\r\n\r\nThe method was just to take the predicted output from my NeMO Quartz15x15 trained ASR net, pair it with the gold translation, and train a new net. The new net was this Seq2Seq model from Ben Trevett: https://github.com/bentrevett/pytorch-seq2seq/blob/master/6%20-%20Attention%20is%20All%20You%20Need.ipynb\r\n\r\nI suppose I could have done the equivalent model in NeMo but they [deleted the example](https://github.com/NVIDIA/NeMo/tree/fa68d336cfefbbbb1849ff1b6ef454149f45234d/examples/nlp/asr_postprocessor). For the NIST exercise I was doing, I needed to train from scratch, and the (deleted) example only showed starting from a pretrained model. I already knew the Ben Trevett code so I went with that.\r\n\r\nThe score after adding this \"afterburner\" net trained on the mistakes of the ASR net was only negligeably higher than without it. "
] | 2020-10-30T18:17:27 | 2020-11-11T13:39:20 | 2020-10-30T21:41:44 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | BERT ASR postprocessing is described here: https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/v0.10.1/nlp/asr-improvement.html
Is there a notebook in the GitHub to demo this in combination with transfer learning of QuartzNet15x5 for a new language?
It would be very helpful to show how to transfer-learn the Bert for a transfer-learned new language QuartzNet.
Also...what happened to all the files referenced in the docs, for example `examples/nlp/asr_postprocessor`
After some digging, I found them buried in an old version. I hope this doesn't mean this doesn't work any more:
https://github.com/NVIDIA/NeMo/tree/fa68d336cfefbbbb1849ff1b6ef454149f45234d/examples/nlp/asr_postprocessor | {
"login": "catskillsresearch",
"id": 38082822,
"node_id": "MDQ6VXNlcjM4MDgyODIy",
"avatar_url": "https://avatars.githubusercontent.com/u/38082822?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/catskillsresearch",
"html_url": "https://github.com/catskillsresearch",
"followers_url": "https://api.github.com/users/catskillsresearch/followers",
"following_url": "https://api.github.com/users/catskillsresearch/following{/other_user}",
"gists_url": "https://api.github.com/users/catskillsresearch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/catskillsresearch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/catskillsresearch/subscriptions",
"organizations_url": "https://api.github.com/users/catskillsresearch/orgs",
"repos_url": "https://api.github.com/users/catskillsresearch/repos",
"events_url": "https://api.github.com/users/catskillsresearch/events{/privacy}",
"received_events_url": "https://api.github.com/users/catskillsresearch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1380/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1380/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1379 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1379/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1379/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1379/events | https://github.com/NVIDIA/NeMo/issues/1379 | 732,962,795 | MDU6SXNzdWU3MzI5NjI3OTU= | 1,379 | FileNotFoundError: Make sure that your data follows the standard format supported by JointIntentSlotDataset. | {
"login": "laifuchicago",
"id": 8537280,
"node_id": "MDQ6VXNlcjg1MzcyODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/8537280?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/laifuchicago",
"html_url": "https://github.com/laifuchicago",
"followers_url": "https://api.github.com/users/laifuchicago/followers",
"following_url": "https://api.github.com/users/laifuchicago/following{/other_user}",
"gists_url": "https://api.github.com/users/laifuchicago/gists{/gist_id}",
"starred_url": "https://api.github.com/users/laifuchicago/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/laifuchicago/subscriptions",
"organizations_url": "https://api.github.com/users/laifuchicago/orgs",
"repos_url": "https://api.github.com/users/laifuchicago/repos",
"events_url": "https://api.github.com/users/laifuchicago/events{/privacy}",
"received_events_url": "https://api.github.com/users/laifuchicago/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1485815557,
"node_id": "MDU6TGFiZWwxNDg1ODE1NTU3",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/bug",
"name": "bug",
"color": "d73a4a",
"default": true,
"description": "Something isn't working"
}
] | closed | false | {
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "vladgets",
"id": 8079846,
"node_id": "MDQ6VXNlcjgwNzk4NDY=",
"avatar_url": "https://avatars.githubusercontent.com/u/8079846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vladgets",
"html_url": "https://github.com/vladgets",
"followers_url": "https://api.github.com/users/vladgets/followers",
"following_url": "https://api.github.com/users/vladgets/following{/other_user}",
"gists_url": "https://api.github.com/users/vladgets/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vladgets/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vladgets/subscriptions",
"organizations_url": "https://api.github.com/users/vladgets/orgs",
"repos_url": "https://api.github.com/users/vladgets/repos",
"events_url": "https://api.github.com/users/vladgets/events{/privacy}",
"received_events_url": "https://api.github.com/users/vladgets/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [] | 2020-10-30T07:56:36 | 2021-06-10T19:03:24 | 2021-06-10T19:03:24 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | To Author:
Now I am using your intent_slot_classification.py. The model could be trained, however, the FileNotFoundError was shown. I just discovered you add the inference part.
Jonathan Sung | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1379/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1379/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/1378 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1378/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1378/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1378/events | https://github.com/NVIDIA/NeMo/pull/1378 | 732,908,862 | MDExOlB1bGxSZXF1ZXN0NTEyNzc5MTgy | 1,378 | PunctuationCapitalizationModel exportable | {
"login": "drnikolaev",
"id": 7504104,
"node_id": "MDQ6VXNlcjc1MDQxMDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7504104?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/drnikolaev",
"html_url": "https://github.com/drnikolaev",
"followers_url": "https://api.github.com/users/drnikolaev/followers",
"following_url": "https://api.github.com/users/drnikolaev/following{/other_user}",
"gists_url": "https://api.github.com/users/drnikolaev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/drnikolaev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/drnikolaev/subscriptions",
"organizations_url": "https://api.github.com/users/drnikolaev/orgs",
"repos_url": "https://api.github.com/users/drnikolaev/repos",
"events_url": "https://api.github.com/users/drnikolaev/events{/privacy}",
"received_events_url": "https://api.github.com/users/drnikolaev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2020-10-30T06:04:42 | 2020-10-31T00:04:56 | 2020-10-31T00:04:56 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1378",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1378",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1378.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1378.patch",
"merged_at": "2020-10-31T00:04:56"
} | Signed-off-by: Sergei Nikolaev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1378/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1378/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1377 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1377/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1377/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1377/events | https://github.com/NVIDIA/NeMo/pull/1377 | 732,876,195 | MDExOlB1bGxSZXF1ZXN0NTEyNzUyMjc0 | 1,377 | Fix http 500 ngc v1.0.0b2 | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging 4f0baa600e628c3bc964be157a258d7dea590259 into 0c575303f49dc0060ad1ff13f3ec9f6e2dcbc381 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-d58640ed1f19ca480d854b6e6914fa40c21d5cc6)\n\n**new alerts:**\n\n* 1 for Except block handles 'BaseException'",
"This pull request **introduces 1 alert** when merging 87b10c1890a5170bc752316be41e9b6d1dc5c9c0 into 0c575303f49dc0060ad1ff13f3ec9f6e2dcbc381 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-07e82da68288ebddb7b2d77a76ed3fc2850929d0)\n\n**new alerts:**\n\n* 1 for Except block handles 'BaseException'",
"looks like it fails on code formatting though",
"This pull request **introduces 1 alert** when merging cb0a5176876f80cacb571f0e761535b2bdb90f20 into 0c575303f49dc0060ad1ff13f3ec9f6e2dcbc381 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-cefe62200e4bdc5bc858569e3d4ff3c0ac446297)\n\n**new alerts:**\n\n* 1 for Except block handles 'BaseException'"
] | 2020-10-30T04:38:14 | 2020-10-30T14:14:17 | 2020-10-30T14:14:11 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/1377",
"html_url": "https://github.com/NVIDIA/NeMo/pull/1377",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/1377.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/1377.patch",
"merged_at": "2020-10-30T14:14:11"
} | Try and wait in case NGC download fails. | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1377/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1377/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/1376 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/1376/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/1376/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/1376/events | https://github.com/NVIDIA/NeMo/issues/1376 | 732,870,252 | MDU6SXNzdWU3MzI4NzAyNTI= | 1,376 | How do I turn off logging? | {
"login": "catskillsresearch",
"id": 38082822,
"node_id": "MDQ6VXNlcjM4MDgyODIy",
"avatar_url": "https://avatars.githubusercontent.com/u/38082822?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/catskillsresearch",
"html_url": "https://github.com/catskillsresearch",
"followers_url": "https://api.github.com/users/catskillsresearch/followers",
"following_url": "https://api.github.com/users/catskillsresearch/following{/other_user}",
"gists_url": "https://api.github.com/users/catskillsresearch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/catskillsresearch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/catskillsresearch/subscriptions",
"organizations_url": "https://api.github.com/users/catskillsresearch/orgs",
"repos_url": "https://api.github.com/users/catskillsresearch/repos",
"events_url": "https://api.github.com/users/catskillsresearch/events{/privacy}",
"received_events_url": "https://api.github.com/users/catskillsresearch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This is my question too.",
"Try this after importing nemo modules:\r\n```python\r\nimport logging\r\nlogging.getLogger('nemo_logger').setLevel(logging.ERROR)\r\n```\r\nyou can choose the logging level from https://docs.python.org/3/library/logging.html#levels",
"[This](https://stackoverflow.com/questions/27647077/fully-disable-python-logging) does the trick, prior to importing anything from NeMo:\r\n\r\n```\r\nimport logging\r\nlogging.disable(logging.CRITICAL)\r\nimport nemo.collections.asr\r\n```\r\nThis won't get rid of some noise about Kaldi:\r\n\r\n```\r\n################################################################################\r\n### WARNING, path does not exist: KALDI_ROOT=/tmp/kaldi\r\n### (please add 'export KALDI_ROOT=<your_path>' in your $HOME/.profile)\r\n### (or run as: KALDI_ROOT=<your_path> python <your_script>.py)\r\n################################################################################\r\n```\r\n\r\nTo get rid of the Kaldi warning, download Kaldi and set KALDI_ROOT. So something like this prior to running the above script (you can put `kaldi` anywere, it doesn't have to be `/tmp`):\r\n\r\n```\r\ncd /tmp\r\ngit clone https://github.com/kaldi-asr/kaldi.git\r\nexport KALDI_ROOT=/tmp/kaldi\r\n```\r\n",
"@burgil the initial question was posed by me for my needs. Turning off all logging output satisfies my original question. How do you see that as \"poor\" when I am the person asking and this method does exactly what I wanted it to? Poor for who?\r\n\r\nAnother way of phrasing your comment would be to say \"I see how this works for you, but for my needs that would cut out too much information, so for my application I would prefer to omit critical\".",
"If you want to disable only NeMo loggings and not all of them, just use this:\r\n\r\n```\r\nlogging.basicConfig(level = logging.INFO)\r\nlogging.getLogger(\"nemo_logger\").setLevel(logging.ERROR)\r\n```"
] | 2020-10-30T04:21:34 | 2023-10-24T13:10:22 | 2021-05-22T08:41:58 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | How do I turn off all the logging messages in NeMo? For example all of this:
```
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text.AudioToCharDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text.AudioToBPEDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text.AudioLabelDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text._TarredAudioToTextDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text.TarredAudioToCharDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2020-10-30 00:16:51 experimental:28] Module <class 'nemo.collections.asr.data.audio_to_text.TarredAudioToBPEDataset'> is experimental, not ready for production and is not fully supported. Use at your own risk.
wandb: WARNING W&B installed but not logged in. Run `wandb login` or set the WANDB_API_KEY env variable.
[NeMo W 2020-10-30 00:16:52 experimental:28] Module <class 'nemo.collections.asr.losses.ctc.CTCLoss'> is experimental, not ready for production and is not fully supported. Use at your own risk.
################################################################################
### WARNING, path does not exist: KALDI_ROOT=/mnt/matylda5/iveselyk/Tools/kaldi-trunk
### (please add 'export KALDI_ROOT=<your_path>' in your $HOME/.profile)
### (or run as: KALDI_ROOT=<your_path> python <your_script>.py)
################################################################################
``` | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/1376/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/NVIDIA/NeMo/issues/1376/timeline | null | completed | false |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.