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 listlengths 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/7952 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7952/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7952/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7952/events | https://github.com/NVIDIA/NeMo/pull/7952 | 2,016,669,715 | PR_kwDOC_bI7s5grHWM | 7,952 | Unfinished checkpoints handling | {
"login": "jbieniusiewi",
"id": 152396322,
"node_id": "U_kgDOCRViIg",
"avatar_url": "https://avatars.githubusercontent.com/u/152396322?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbieniusiewi",
"html_url": "https://github.com/jbieniusiewi",
"followers_url": "https://api.github.com/users/jbieniusiewi/followers",
"following_url": "https://api.github.com/users/jbieniusiewi/following{/other_user}",
"gists_url": "https://api.github.com/users/jbieniusiewi/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbieniusiewi/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbieniusiewi/subscriptions",
"organizations_url": "https://api.github.com/users/jbieniusiewi/orgs",
"repos_url": "https://api.github.com/users/jbieniusiewi/repos",
"events_url": "https://api.github.com/users/jbieniusiewi/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbieniusiewi/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
}
] | closed | false | null | [] | null | [
"jenkins",
"Hi @titu1994,\r\n\r\nThank you for the review! \r\nI added the docstrings and comments.\r\n\r\nRegarding your question: If all checkpoints gets filtered out in Exp Manager, then depending on `resume_ignore_no_checkpoint` option - training starts from scratch (resume_ignore_no_checkpoint=True) or rais... | 2023-11-29T14:04:53 | 2024-02-08T06:12:31 | 2024-02-08T06:11:27 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7952",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7952",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7952.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7952.patch",
"merged_at": "2024-02-08T06:11:27"
} | # What does this PR do ?
Handle incomplete checkpoints.
With this fix, if last checkpointing operation was interrupted for some reason and left incomplete data, training should be resumed from the last available full checkpoint, instead of trying to read the incomplete checkpoint and crashing. This should simplify training auto-resume logic.
Implementation overview:
Create "unfinished" marker file for a checkpoint that is being written or removed.
The marker file is deleted after the checkpoint IO operation completes.
When resuming, checkpoints marked as unfinished are ignored and subsequently removed.
**Collection**:
Core
# Before your PR is "Ready for review"
**Pre checks**:
- [X] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [X] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [X] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
@ericharper, @titu1994, @blisc, @okuchaiev | {
"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/7952/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/7952/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7951 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7951/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7951/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7951/events | https://github.com/NVIDIA/NeMo/pull/7951 | 2,015,516,818 | PR_kwDOC_bI7s5gnL6e | 7,951 | Enable pipelined tp comm overlap for mcore path | {
"login": "erhoo82",
"id": 15861259,
"node_id": "MDQ6VXNlcjE1ODYxMjU5",
"avatar_url": "https://avatars.githubusercontent.com/u/15861259?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/erhoo82",
"html_url": "https://github.com/erhoo82",
"followers_url": "https://api.github.com/users/erhoo82/followers",
"following_url": "https://api.github.com/users/erhoo82/following{/other_user}",
"gists_url": "https://api.github.com/users/erhoo82/gists{/gist_id}",
"starred_url": "https://api.github.com/users/erhoo82/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/erhoo82/subscriptions",
"organizations_url": "https://api.github.com/users/erhoo82/orgs",
"repos_url": "https://api.github.com/users/erhoo82/repos",
"events_url": "https://api.github.com/users/erhoo82/events{/privacy}",
"received_events_url": "https://api.github.com/users/erhoo82/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"MCore code path ignores config of enabling or disabling TE, so this LGTM. Thanks.",
"jenkins",
"jenkins",
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been... | 2023-11-28T23:47:12 | 2023-12-22T01:45:08 | 2023-12-22T01:45:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7951",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7951",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7951.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7951.patch",
"merged_at": null
} | # What does this PR do ?
Enable pipelined tp comm overlap for mcore path
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Jenkins CI
To run Jenkins, a NeMo User with write access must comment `jenkins` on the PR.
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7951/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/7951/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7950 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7950/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7950/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7950/events | https://github.com/NVIDIA/NeMo/pull/7950 | 2,014,674,577 | PR_kwDOC_bI7s5gkSNH | 7,950 | Draft: SFT Jinja Templating | {
"login": "aleksficek",
"id": 37374704,
"node_id": "MDQ6VXNlcjM3Mzc0NzA0",
"avatar_url": "https://avatars.githubusercontent.com/u/37374704?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/aleksficek",
"html_url": "https://github.com/aleksficek",
"followers_url": "https://api.github.com/users/aleksficek/followers",
"following_url": "https://api.github.com/users/aleksficek/following{/other_user}",
"gists_url": "https://api.github.com/users/aleksficek/gists{/gist_id}",
"starred_url": "https://api.github.com/users/aleksficek/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/aleksficek/subscriptions",
"organizations_url": "https://api.github.com/users/aleksficek/orgs",
"repos_url": "https://api.github.com/users/aleksficek/repos",
"events_url": "https://api.github.com/users/aleksficek/events{/privacy}",
"received_events_url": "https://api.github.com/users/aleksficek/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was c... | 2023-11-28T15:14:29 | 2024-01-09T01:46:19 | 2024-01-09T01:46:18 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7950",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7950",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7950.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7950.patch",
"merged_at": null
} | # What does this PR do ?
Adds support for Jinja2 templates for input prompt and chat prompts.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7950/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/7950/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7949 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7949/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7949/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7949/events | https://github.com/NVIDIA/NeMo/issues/7949 | 2,013,708,870 | I_kwDOC_bI7s54BsJG | 7,949 | The training time for SpeakerNet is too long | {
"login": "DTDwind",
"id": 22238266,
"node_id": "MDQ6VXNlcjIyMjM4MjY2",
"avatar_url": "https://avatars.githubusercontent.com/u/22238266?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/DTDwind",
"html_url": "https://github.com/DTDwind",
"followers_url": "https://api.github.com/users/DTDwind/followers",
"following_url": "https://api.github.com/users/DTDwind/following{/other_user}",
"gists_url": "https://api.github.com/users/DTDwind/gists{/gist_id}",
"starred_url": "https://api.github.com/users/DTDwind/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/DTDwind/subscriptions",
"organizations_url": "https://api.github.com/users/DTDwind/orgs",
"repos_url": "https://api.github.com/users/DTDwind/repos",
"events_url": "https://api.github.com/users/DTDwind/events{/privacy}",
"received_events_url": "https://api.github.com/users/DTDwind/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 | [
"\r\nI found that setting `num_sanity_val_steps: 0` and` limit_val_batches: 0.0` allows the training stage to run without the `Sanity Checking DataLoader\"`and `Validation DataLoader`. However, I still want to know if there is a better way to handle this, as I believe a time requirement of 66452 hours is not a norm... | 2023-11-28T05:57:05 | 2024-01-09T05:19:15 | 2024-01-09T05:19:15 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi,
I am training a SpeakerNet, and my process is as follows:
Step 1: In the `SpeakerNet_verification_3x2x256.yaml` file,
- Set `train_ds manifest_filepath` to `path/to/my/trainset`.
- Set `validation_ds manifest_filepath` to `path/to/my/trainset`.
It's worth noting that `train_ds` and `validation_ds` have the same `manifest_filepath` because I found that not doing so results in errors.
Step 2: Execute `speaker_reco.py`.
I noticed that the program spends a considerable amount of time in the `Sanity Checking DataLoader` and also in the `Validation DataLoader` stage. The program estimates that it will take 66452 hours to complete one epoch. I believe there is some issue, and I'm wondering if anyone can provide me with advice on this matter.
Tool version
python 3.10.13
NeMo 1.20.0
pytorch-lightning 1.9.4
torch 2.1.0
| {
"login": "DTDwind",
"id": 22238266,
"node_id": "MDQ6VXNlcjIyMjM4MjY2",
"avatar_url": "https://avatars.githubusercontent.com/u/22238266?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/DTDwind",
"html_url": "https://github.com/DTDwind",
"followers_url": "https://api.github.com/users/DTDwind/followers",
"following_url": "https://api.github.com/users/DTDwind/following{/other_user}",
"gists_url": "https://api.github.com/users/DTDwind/gists{/gist_id}",
"starred_url": "https://api.github.com/users/DTDwind/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/DTDwind/subscriptions",
"organizations_url": "https://api.github.com/users/DTDwind/orgs",
"repos_url": "https://api.github.com/users/DTDwind/repos",
"events_url": "https://api.github.com/users/DTDwind/events{/privacy}",
"received_events_url": "https://api.github.com/users/DTDwind/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7949/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/7949/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7948 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7948/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7948/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7948/events | https://github.com/NVIDIA/NeMo/issues/7948 | 2,013,368,989 | I_kwDOC_bI7s54AZKd | 7,948 | Multiple last checkpoints | {
"login": "szhengac",
"id": 3960020,
"node_id": "MDQ6VXNlcjM5NjAwMjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/3960020?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/szhengac",
"html_url": "https://github.com/szhengac",
"followers_url": "https://api.github.com/users/szhengac/followers",
"following_url": "https://api.github.com/users/szhengac/following{/other_user}",
"gists_url": "https://api.github.com/users/szhengac/gists{/gist_id}",
"starred_url": "https://api.github.com/users/szhengac/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/szhengac/subscriptions",
"organizations_url": "https://api.github.com/users/szhengac/orgs",
"repos_url": "https://api.github.com/users/szhengac/repos",
"events_url": "https://api.github.com/users/szhengac/events{/privacy}",
"received_events_url": "https://api.github.com/users/szhengac/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 | [
"This happens when the job is killed while saving a checkpoint. Currently you need to manually cleanup the checkpoints folder to be able to resume. Figure out which is the most recent *complete* \"-last\" checkpoint, and delete the other(s).",
"pls re-open if solution doesn;t work or if you face any other issue. ... | 2023-11-28T00:30:27 | 2024-01-08T19:46:26 | 2023-12-23T18:17:04 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Received the following error when I killed the job and try to resume it. It uses mcore gpt.
```
Traceback (most recent call last):
File "/opt/NeMo/examples/nlp/language_modeling/megatron_gpt_pretraining.py", line 34, in main
exp_manager(trainer, cfg.exp_manager)
File "/opt/NeMo/nemo/utils/exp_manager.py", line 348, in exp_manager
check_resume(
File "/opt/NeMo/nemo/utils/exp_manager.py", line 622, in check_resume
raise ValueError(f"Multiple checkpoints {last_checkpoints} that matches *last.ckpt.")
ValueError: Multiple checkpoints [PosixPath('/results/checkpoints/boson-consumed_samples=6224.0-last'), PosixPath('/results/checkpoints/boson-consumed_samples=4800.0-last')] that matches *last.ckpt.
``` | {
"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/7948/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/7948/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7947 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7947/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7947/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7947/events | https://github.com/NVIDIA/NeMo/pull/7947 | 2,013,250,735 | PR_kwDOC_bI7s5gfbMo | 7,947 | [TTS] Scale sampler steps by number of devices | {
"login": "rlangman",
"id": 9942053,
"node_id": "MDQ6VXNlcjk5NDIwNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/9942053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rlangman",
"html_url": "https://github.com/rlangman",
"followers_url": "https://api.github.com/users/rlangman/followers",
"following_url": "https://api.github.com/users/rlangman/following{/other_user}",
"gists_url": "https://api.github.com/users/rlangman/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rlangman/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rlangman/subscriptions",
"organizations_url": "https://api.github.com/users/rlangman/orgs",
"repos_url": "https://api.github.com/users/rlangman/repos",
"events_url": "https://api.github.com/users/rlangman/events{/privacy}",
"received_events_url": "https://api.github.com/users/rlangman/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"jenkins"
] | 2023-11-27T22:40:41 | 2023-12-13T19:04:37 | 2023-12-13T19:04:36 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7947",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7947",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7947.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7947.patch",
"merged_at": "2023-12-13T19:04:36"
} | # What does this PR do ?
The weighted random sampler is configured to produce a fixed number of training samples during each epoch. This means if you use multiple GPUs, then the number of steps in each epoch is scaled down by the number of GPUs. This change multiplies the number of samples by the world size to keep the step count independent of number of GPUs used.
**Collection**: [TTS]
# Changelog
- Scale number of samples per epoch by world size when using weighted random sampler.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
| {
"login": "rlangman",
"id": 9942053,
"node_id": "MDQ6VXNlcjk5NDIwNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/9942053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rlangman",
"html_url": "https://github.com/rlangman",
"followers_url": "https://api.github.com/users/rlangman/followers",
"following_url": "https://api.github.com/users/rlangman/following{/other_user}",
"gists_url": "https://api.github.com/users/rlangman/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rlangman/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rlangman/subscriptions",
"organizations_url": "https://api.github.com/users/rlangman/orgs",
"repos_url": "https://api.github.com/users/rlangman/repos",
"events_url": "https://api.github.com/users/rlangman/events{/privacy}",
"received_events_url": "https://api.github.com/users/rlangman/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7947/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/7947/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7946 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7946/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7946/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7946/events | https://github.com/NVIDIA/NeMo/pull/7946 | 2,013,202,726 | PR_kwDOC_bI7s5gfQZk | 7,946 | only enable query key scaling during fp16 | {
"login": "gshennvm",
"id": 119401249,
"node_id": "U_kgDOBx3rIQ",
"avatar_url": "https://avatars.githubusercontent.com/u/119401249?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gshennvm",
"html_url": "https://github.com/gshennvm",
"followers_url": "https://api.github.com/users/gshennvm/followers",
"following_url": "https://api.github.com/users/gshennvm/following{/other_user}",
"gists_url": "https://api.github.com/users/gshennvm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gshennvm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gshennvm/subscriptions",
"organizations_url": "https://api.github.com/users/gshennvm/orgs",
"repos_url": "https://api.github.com/users/gshennvm/repos",
"events_url": "https://api.github.com/users/gshennvm/events{/privacy}",
"received_events_url": "https://api.github.com/users/gshennvm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-27T22:11:17 | 2023-12-01T20:30:02 | 2023-12-01T20:30:01 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7946",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7946",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7946.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7946.patch",
"merged_at": "2023-12-01T20:30:01"
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
Enable query key scaling only during fp16, since this is when TE enables query key scaling. https://github.com/NVIDIA/TransformerEngine/blob/666539f36275fa9c0fbc99f9ea50f2d6e29e336f/transformer_engine/pytorch/attention.py#L940
| {
"login": "gshennvm",
"id": 119401249,
"node_id": "U_kgDOBx3rIQ",
"avatar_url": "https://avatars.githubusercontent.com/u/119401249?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gshennvm",
"html_url": "https://github.com/gshennvm",
"followers_url": "https://api.github.com/users/gshennvm/followers",
"following_url": "https://api.github.com/users/gshennvm/following{/other_user}",
"gists_url": "https://api.github.com/users/gshennvm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gshennvm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gshennvm/subscriptions",
"organizations_url": "https://api.github.com/users/gshennvm/orgs",
"repos_url": "https://api.github.com/users/gshennvm/repos",
"events_url": "https://api.github.com/users/gshennvm/events{/privacy}",
"received_events_url": "https://api.github.com/users/gshennvm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7946/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/7946/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7945 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7945/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7945/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7945/events | https://github.com/NVIDIA/NeMo/pull/7945 | 2,013,184,966 | PR_kwDOC_bI7s5gfMYg | 7,945 | Packed Sequence | {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-27T22:01:12 | 2024-03-16T05:10:52 | 2023-12-16T01:25:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7945",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7945",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7945.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7945.patch",
"merged_at": "2023-12-16T01:25:59"
} | # What does this PR do ?
Support training with packed sequences for SFT and PEFT.
In this scenario, sequences in a batch are "concatenated" in the sequence length dimension, instead of "stacked" in a batch dimension.
This is shown to significantly improve the training throughput.
Packed sequences will need to be prepared in a particular way (script in #8682 )
**Collection**: NLP
# Changelog
- new GPTSFTPackedDataset class to handle the new dataset format
- pass `cu_seqlens` and `qkv_format` to the forward() function of mcore model
# Usage
Note: only train_ds is packed. Validation and test datasets are unchanged.
```python
python /opt/NeMo/examples/nlp/language_modeling/tuning/megatron_gpt_peft_tuning.py \
...
+model.data.train_ds.packed_sequence=True
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7945/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/7945/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7944 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7944/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7944/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7944/events | https://github.com/NVIDIA/NeMo/pull/7944 | 2,012,983,275 | PR_kwDOC_bI7s5gef1g | 7,944 | Instructions for running ci on pr template | {
"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
} | [
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/CI",
"name": "CI",
"color": "1D76DB",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-27T19:48:14 | 2023-11-28T16:34:11 | 2023-11-28T16:34:10 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7944",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7944",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7944.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7944.patch",
"merged_at": "2023-11-28T16:34:10"
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7944/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/7944/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7943 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7943/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7943/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7943/events | https://github.com/NVIDIA/NeMo/pull/7943 | 2,012,865,737 | PR_kwDOC_bI7s5geF4B | 7,943 | Add checks for unit tests that are looking for data from CI machine | {
"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
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"tested and it looks good to me"
] | 2023-11-27T18:35:08 | 2023-11-28T00:38:08 | 2023-11-28T00:38:07 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7943",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7943",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7943.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7943.patch",
"merged_at": "2023-11-28T00:38:07"
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7943/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/7943/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7942 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7942/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7942/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7942/events | https://github.com/NVIDIA/NeMo/pull/7942 | 2,012,755,755 | PR_kwDOC_bI7s5gdt6J | 7,942 | Added a procedure for Windows users, README | {
"login": "Jorjeous",
"id": 37293288,
"node_id": "MDQ6VXNlcjM3MjkzMjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/37293288?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Jorjeous",
"html_url": "https://github.com/Jorjeous",
"followers_url": "https://api.github.com/users/Jorjeous/followers",
"following_url": "https://api.github.com/users/Jorjeous/following{/other_user}",
"gists_url": "https://api.github.com/users/Jorjeous/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Jorjeous/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Jorjeous/subscriptions",
"organizations_url": "https://api.github.com/users/Jorjeous/orgs",
"repos_url": "https://api.github.com/users/Jorjeous/repos",
"events_url": "https://api.github.com/users/Jorjeous/events{/privacy}",
"received_events_url": "https://api.github.com/users/Jorjeous/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1485815558,
"node_id": "MDU6TGFiZWwxNDg1ODE1NTU4",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/documentation",
"name": "documentation",
"color": "0075ca",
"default": true,
"description": "Improvements or additions to documentation"
}
] | closed | false | {
"login": "Jorjeous",
"id": 37293288,
"node_id": "MDQ6VXNlcjM3MjkzMjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/37293288?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Jorjeous",
"html_url": "https://github.com/Jorjeous",
"followers_url": "https://api.github.com/users/Jorjeous/followers",
"following_url": "https://api.github.com/users/Jorjeous/following{/other_user}",
"gists_url": "https://api.github.com/users/Jorjeous/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Jorjeous/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Jorjeous/subscriptions",
"organizations_url": "https://api.github.com/users/Jorjeous/orgs",
"repos_url": "https://api.github.com/users/Jorjeous/repos",
"events_url": "https://api.github.com/users/Jorjeous/events{/privacy}",
"received_events_url": "https://api.github.com/users/Jorjeous/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "Jorjeous",
"id": 37293288,
"node_id": "MDQ6VXNlcjM3MjkzMjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/37293288?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Jorjeous",
"html_url": "https://github.com/Jorjeous",
"followers_url": "https://api.... | null | [
"[celebrate] Georgy Zelenfroind reacted to your message:\r\n________________________________\r\nFrom: Somshubra Majumdar ***@***.***>\r\nSent: Tuesday, December 12, 2023 3:54:39 PM\r\nTo: NVIDIA/NeMo ***@***.***>\r\nCc: Georgy Zelenfroind ***@***.***>; Assign ***@***.***>\r\nSubject: Re: [NVIDIA/NeMo] Added a p... | 2023-11-27T17:26:07 | 2023-12-13T18:05:39 | 2023-12-12T15:54:40 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7942",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7942",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7942.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7942.patch",
"merged_at": "2023-12-12T15:54:40"
} | Added a brief guide and official links for troubleshooting WSL installation for Windows users
# What does this PR do ?
Adds instruction for Windows users
**Collection**: [Note which collection this PR will affect]
Documentation
# Changelog
Added text
# Usage
Reading
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [-] Did you write any new necessary tests?
- [+] Did you add or update any necessary documentation?
- [-] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [+] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7942/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/7942/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7941 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7941/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7941/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7941/events | https://github.com/NVIDIA/NeMo/pull/7941 | 2,012,460,507 | PR_kwDOC_bI7s5gctPY | 7,941 | Fix HF dataset streaming conversion | {
"login": "bofenghuang",
"id": 38185248,
"node_id": "MDQ6VXNlcjM4MTg1MjQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/38185248?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bofenghuang",
"html_url": "https://github.com/bofenghuang",
"followers_url": "https://api.github.com/users/bofenghuang/followers",
"following_url": "https://api.github.com/users/bofenghuang/following{/other_user}",
"gists_url": "https://api.github.com/users/bofenghuang/gists{/gist_id}",
"starred_url": "https://api.github.com/users/bofenghuang/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/bofenghuang/subscriptions",
"organizations_url": "https://api.github.com/users/bofenghuang/orgs",
"repos_url": "https://api.github.com/users/bofenghuang/repos",
"events_url": "https://api.github.com/users/bofenghuang/events{/privacy}",
"received_events_url": "https://api.github.com/users/bofenghuang/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-27T14:56:23 | 2023-12-20T01:39:55 | 2023-12-20T01:39:54 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7941",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7941",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7941.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7941.patch",
"merged_at": null
} | # What does this PR do ?
Hi 👋, tiny modif here to save the right item into the manifest.
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7941/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/7941/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7940 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7940/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7940/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7940/events | https://github.com/NVIDIA/NeMo/pull/7940 | 2,012,453,225 | PR_kwDOC_bI7s5gcrrE | 7,940 | SDE2: GPU acceleration | {
"login": "ssh-meister",
"id": 117230141,
"node_id": "U_kgDOBvzKPQ",
"avatar_url": "https://avatars.githubusercontent.com/u/117230141?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ssh-meister",
"html_url": "https://github.com/ssh-meister",
"followers_url": "https://api.github.com/users/ssh-meister/followers",
"following_url": "https://api.github.com/users/ssh-meister/following{/other_user}",
"gists_url": "https://api.github.com/users/ssh-meister/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ssh-meister/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ssh-meister/subscriptions",
"organizations_url": "https://api.github.com/users/ssh-meister/orgs",
"repos_url": "https://api.github.com/users/ssh-meister/repos",
"events_url": "https://api.github.com/users/ssh-meister/events{/privacy}",
"received_events_url": "https://api.github.com/users/ssh-meister/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | 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/use... | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-27T14:52:46 | 2024-01-19T01:46:56 | 2024-01-19T01:46:56 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7940",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7940",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7940.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7940.patch",
"merged_at": null
} | GPU acceleration added,
Data loading optimization with adding multiprocessing,
Code refactoring | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7940/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/7940/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7939 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7939/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7939/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7939/events | https://github.com/NVIDIA/NeMo/issues/7939 | 2,011,924,400 | I_kwDOC_bI7s5364ew | 7,939 | Nemotron-3-8b finetuning | {
"login": "LeonardoEmili",
"id": 36575651,
"node_id": "MDQ6VXNlcjM2NTc1NjUx",
"avatar_url": "https://avatars.githubusercontent.com/u/36575651?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LeonardoEmili",
"html_url": "https://github.com/LeonardoEmili",
"followers_url": "https://api.github.com/users/LeonardoEmili/followers",
"following_url": "https://api.github.com/users/LeonardoEmili/following{/other_user}",
"gists_url": "https://api.github.com/users/LeonardoEmili/gists{/gist_id}",
"starred_url": "https://api.github.com/users/LeonardoEmili/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/LeonardoEmili/subscriptions",
"organizations_url": "https://api.github.com/users/LeonardoEmili/orgs",
"repos_url": "https://api.github.com/users/LeonardoEmili/repos",
"events_url": "https://api.github.com/users/LeonardoEmili/events{/privacy}",
"received_events_url": "https://api.github.com/users/LeonardoEmili/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"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
} | [
{
"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:... | null | [
"Is `nemofw-training:23.08.03` subset of `nvcr.io/nvidia/nemo:23.08`? \r\n\r\nThe official guide [here](https://docs.nvidia.com/nemo-framework/user-guide/latest/playbooks/llama2sft.html) suggests to use `nemofw-training:23.08.03`, but I got the following error when I `docker pull nvcr.io/ea-bignlp/ga-participants/n... | 2023-11-27T10:06:30 | 2024-03-02T06:28:05 | 2024-01-10T10:14:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I was looking into this [newly released model](https://developer.nvidia.com/blog/nvidia-ai-foundation-models-build-custom-enterprise-chatbots-and-co-pilots-with-production-ready-llms/) and wanted to experiment with either SFT or PEFT-finetuning but it seems documentation is lacking on it (the [only tutorial](https://huggingface.co/nvidia/nemotron-3-8b-base-4k) available seems to be on inference at the moment). Actually, I couldn't find any mention to the model in the code so I'm not even sure training is possible at all with code available.
In support to above, I've tried the code using the latest available image (`nvcr.io/nvidia/nemo:23.08`) with a small-sized model (gpt-2) and it works fine. However, when plugging in the nemotron weights it raises an exception at [L1221](https://github.com/NVIDIA/TransformerEngine/blame/db589510a69b5b5c655d439ad3521ea890c7b454/transformer_engine/pytorch/attention.py#L1221) due to an unsupported Q,K,V layout. Note that this code no longer exist in main after [this PR](https://github.com/NVIDIA/TransformerEngine/pull/403).
Could you please confirm that the code available in this public repo _supports nemotron finetuning_ **and** provide some guidance on how to do it (e.g. pointer to the correct docker image, example usage would also be great)?
Tagging @ericharper and @nithinraok hoping they can help, thank you in advance! | {
"login": "LeonardoEmili",
"id": 36575651,
"node_id": "MDQ6VXNlcjM2NTc1NjUx",
"avatar_url": "https://avatars.githubusercontent.com/u/36575651?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LeonardoEmili",
"html_url": "https://github.com/LeonardoEmili",
"followers_url": "https://api.github.com/users/LeonardoEmili/followers",
"following_url": "https://api.github.com/users/LeonardoEmili/following{/other_user}",
"gists_url": "https://api.github.com/users/LeonardoEmili/gists{/gist_id}",
"starred_url": "https://api.github.com/users/LeonardoEmili/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/LeonardoEmili/subscriptions",
"organizations_url": "https://api.github.com/users/LeonardoEmili/orgs",
"repos_url": "https://api.github.com/users/LeonardoEmili/repos",
"events_url": "https://api.github.com/users/LeonardoEmili/events{/privacy}",
"received_events_url": "https://api.github.com/users/LeonardoEmili/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7939/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/7939/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7938 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7938/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7938/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7938/events | https://github.com/NVIDIA/NeMo/issues/7938 | 2,010,807,605 | I_kwDOC_bI7s532n01 | 7,938 | [ASR] multi-node multi-gpu training hangs on end of epoch, post validation and checkpoint save | {
"login": "itzsimpl",
"id": 4784070,
"node_id": "MDQ6VXNlcjQ3ODQwNzA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4784070?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/itzsimpl",
"html_url": "https://github.com/itzsimpl",
"followers_url": "https://api.github.com/users/itzsimpl/followers",
"following_url": "https://api.github.com/users/itzsimpl/following{/other_user}",
"gists_url": "https://api.github.com/users/itzsimpl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/itzsimpl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/itzsimpl/subscriptions",
"organizations_url": "https://api.github.com/users/itzsimpl/orgs",
"repos_url": "https://api.github.com/users/itzsimpl/repos",
"events_url": "https://api.github.com/users/itzsimpl/events{/privacy}",
"received_events_url": "https://api.github.com/users/itzsimpl/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"@KunalDhawan @nithinraok Have you seen anything like this?",
"I had something similar, not sure if exactly the same though.\r\n\r\nWhat I did was: \r\n1. Not use the container: Install everything manually inside a virtualenv.\r\n2. Do not use `dali`. \r\n3. Check the callbacks you are using. \r\n",
"The cluste... | 2023-11-25T21:52:46 | 2024-03-09T01:42:29 | 2024-03-09T01:42:29 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I'm running a standard ASR training recipe (conformer ctc) in a multi-node multi-gpu setting using SLURM. The training stops on epoch end, post validation, when the checkpoint is saved. The next epoch never starts. Single-node multi-gpu training runs fine.
NCCL does not seem to be the problem (a simple dummy training based on MNIST runs fine, and the checkpoints are saved). I've also tried running the NeMo ASR code with `gloo` as backend, but there's no change, training hangs post validation after the model is saved.
The shared storage, does not seem to be the issue, either, as single-node trainings work fine, so does training independently, but concurrently, on all nodes. Also, if I disable checkpointing training continues normally till the last epoch, when the end of training model is saved.
The only clue I have is that I'm running using the `nemo:23.08` container, and I've noticed the following log
```python
[NeMo W 2023-11-25 21:45:41 nemo_logging:349] /usr/local/lib/python3.10/dist-packages/pytorch_lightning/trainer/connectors/logger_connector/result.py:433: PossibleUserWarning: It is recommended to use `self.log('global_step', ..., sync_dist=True)` when logging on epoch level in distributed setting to accumulate the metric across devices.
```
Any ideas how to resolve this issue?
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7938/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/7938/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7937 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7937/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7937/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7937/events | https://github.com/NVIDIA/NeMo/issues/7937 | 2,010,682,992 | I_kwDOC_bI7s532JZw | 7,937 | How to apply early stopping criterion while finetuning conformer? | {
"login": "mukherjeesougata",
"id": 95704630,
"node_id": "U_kgDOBbRWNg",
"avatar_url": "https://avatars.githubusercontent.com/u/95704630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mukherjeesougata",
"html_url": "https://github.com/mukherjeesougata",
"followers_url": "https://api.github.com/users/mukherjeesougata/followers",
"following_url": "https://api.github.com/users/mukherjeesougata/following{/other_user}",
"gists_url": "https://api.github.com/users/mukherjeesougata/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mukherjeesougata/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mukherjeesougata/subscriptions",
"organizations_url": "https://api.github.com/users/mukherjeesougata/orgs",
"repos_url": "https://api.github.com/users/mukherjeesougata/repos",
"events_url": "https://api.github.com/users/mukherjeesougata/events{/privacy}",
"received_events_url": "https://api.github.com/users/mukherjeesougata/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "htt... | closed | false | {
"login": "KunalDhawan",
"id": 15515031,
"node_id": "MDQ6VXNlcjE1NTE1MDMx",
"avatar_url": "https://avatars.githubusercontent.com/u/15515031?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/KunalDhawan",
"html_url": "https://github.com/KunalDhawan",
"followers_url": "https://api.github.com/users/KunalDhawan/followers",
"following_url": "https://api.github.com/users/KunalDhawan/following{/other_user}",
"gists_url": "https://api.github.com/users/KunalDhawan/gists{/gist_id}",
"starred_url": "https://api.github.com/users/KunalDhawan/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/KunalDhawan/subscriptions",
"organizations_url": "https://api.github.com/users/KunalDhawan/orgs",
"repos_url": "https://api.github.com/users/KunalDhawan/repos",
"events_url": "https://api.github.com/users/KunalDhawan/events{/privacy}",
"received_events_url": "https://api.github.com/users/KunalDhawan/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "KunalDhawan",
"id": 15515031,
"node_id": "MDQ6VXNlcjE1NTE1MDMx",
"avatar_url": "https://avatars.githubusercontent.com/u/15515031?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/KunalDhawan",
"html_url": "https://github.com/KunalDhawan",
"followers_url": "htt... | null | [
"@okuchaiev A gentle reminder for the same",
"@titu1994 can you kindly please answer this",
"@KunalDhawan could you answer this?",
"Hi @mukherjeesougata, we do provide support for early stopping based on [PTL's EarlyStopping callback](https://lightning.ai/docs/pytorch/stable/common/early_stopping.html). The c... | 2023-11-25T14:39:58 | 2024-01-05T01:45:46 | 2024-01-05T01:45:46 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am trying out conformer-ctc and conformer-transducer finetuning with 5hrs of data using https://storage.googleapis.com/vakyansh-open-models/pretrained_models/vakyansh-conformer-ssl/ssl_conformer_large_e178.nemo pretrained model . I have seen that during training after certain number of epochs the wer is increasing during validation. So, I want to apply early stopping criterion for that. Can anybody kindly please help me out how to do that?
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7937/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/7937/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7936 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7936/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7936/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7936/events | https://github.com/NVIDIA/NeMo/issues/7936 | 2,009,984,808 | I_kwDOC_bI7s53ze8o | 7,936 | How to remove Punctuation and Capitalization from appearing in a P&C ASR model's output? | {
"login": "abarcovschi",
"id": 26311007,
"node_id": "MDQ6VXNlcjI2MzExMDA3",
"avatar_url": "https://avatars.githubusercontent.com/u/26311007?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/abarcovschi",
"html_url": "https://github.com/abarcovschi",
"followers_url": "https://api.github.com/users/abarcovschi/followers",
"following_url": "https://api.github.com/users/abarcovschi/following{/other_user}",
"gists_url": "https://api.github.com/users/abarcovschi/gists{/gist_id}",
"starred_url": "https://api.github.com/users/abarcovschi/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/abarcovschi/subscriptions",
"organizations_url": "https://api.github.com/users/abarcovschi/orgs",
"repos_url": "https://api.github.com/users/abarcovschi/repos",
"events_url": "https://api.github.com/users/abarcovschi/events{/privacy}",
"received_events_url": "https://api.github.com/users/abarcovschi/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | {
"login": "KunalDhawan",
"id": 15515031,
"node_id": "MDQ6VXNlcjE1NTE1MDMx",
"avatar_url": "https://avatars.githubusercontent.com/u/15515031?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/KunalDhawan",
"html_url": "https://github.com/KunalDhawan",
"followers_url": "https://api.github.com/users/KunalDhawan/followers",
"following_url": "https://api.github.com/users/KunalDhawan/following{/other_user}",
"gists_url": "https://api.github.com/users/KunalDhawan/gists{/gist_id}",
"starred_url": "https://api.github.com/users/KunalDhawan/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/KunalDhawan/subscriptions",
"organizations_url": "https://api.github.com/users/KunalDhawan/orgs",
"repos_url": "https://api.github.com/users/KunalDhawan/repos",
"events_url": "https://api.github.com/users/KunalDhawan/events{/privacy}",
"received_events_url": "https://api.github.com/users/KunalDhawan/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "KunalDhawan",
"id": 15515031,
"node_id": "MDQ6VXNlcjE1NTE1MDMx",
"avatar_url": "https://avatars.githubusercontent.com/u/15515031?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/KunalDhawan",
"html_url": "https://github.com/KunalDhawan",
"followers_url": "htt... | null | [
"Hi @abarcovschi, we don't support P&C stripping currently but would be adding this feature very soon! The easiest way would be to filter-out/ convert (capitalization) the tokens in post-processing. We would soon be adding text post-processors for ASR, initial implementation (not tested/merged) is available here - ... | 2023-11-24T16:42:49 | 2023-11-28T00:35:51 | 2023-11-28T00:35:51 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am running `transcribe_speech.py` on the [STT En FastConformer Hybrid Transducer-CTC Large P&C](https://catalog.ngc.nvidia.com/orgs/nvidia/teams/nemo/models/stt_en_fastconformer_hybrid_large_pc) model, but I do not need the P&C feature. Is there a flag I can give to the script to stop the model from generating those tokens, or do I have to filter out the tokens as a postprocessing step myself? | {
"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/7936/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/7936/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7935 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7935/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7935/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7935/events | https://github.com/NVIDIA/NeMo/pull/7935 | 2,009,698,204 | PR_kwDOC_bI7s5gTqLs | 7,935 | [NLP] HF -> NeMo conversion enhancements | {
"login": "janekl",
"id": 5582967,
"node_id": "MDQ6VXNlcjU1ODI5Njc=",
"avatar_url": "https://avatars.githubusercontent.com/u/5582967?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/janekl",
"html_url": "https://github.com/janekl",
"followers_url": "https://api.github.com/users/janekl/followers",
"following_url": "https://api.github.com/users/janekl/following{/other_user}",
"gists_url": "https://api.github.com/users/janekl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/janekl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/janekl/subscriptions",
"organizations_url": "https://api.github.com/users/janekl/orgs",
"repos_url": "https://api.github.com/users/janekl/repos",
"events_url": "https://api.github.com/users/janekl/events{/privacy}",
"received_events_url": "https://api.github.com/users/janekl/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
... | closed | false | null | [] | null | [
"Closing to follow up elsewhere - see the description above for details."
] | 2023-11-24T13:04:14 | 2024-11-19T16:06:45 | 2023-12-06T10:35:47 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7935",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7935",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7935.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7935.patch",
"merged_at": null
} | # Announcement📢📢📢
I'm closing this PR in favor of a series smaller PRs with all the corresponding changes applied.
* https://github.com/NVIDIA/NeMo/pull/7977
* https://github.com/NVIDIA/NeMo/pull/7978
* https://github.com/NVIDIA/NeMo/pull/7979
* https://github.com/NVIDIA/NeMo/pull/7980
* https://github.com/NVIDIA/NeMo/pull/7981
* https://github.com/NVIDIA/NeMo/pull/7982
Hopefully, this makes it easier to review and integrate the changes.
------------------------------------------------------------------------------------------
# What does this PR do ?
This PR improves community models conversion scripts: Llama2, MPT, and StarCoder.
**Collection**: NLP
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [x] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [x] Documentation
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas. | {
"login": "janekl",
"id": 5582967,
"node_id": "MDQ6VXNlcjU1ODI5Njc=",
"avatar_url": "https://avatars.githubusercontent.com/u/5582967?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/janekl",
"html_url": "https://github.com/janekl",
"followers_url": "https://api.github.com/users/janekl/followers",
"following_url": "https://api.github.com/users/janekl/following{/other_user}",
"gists_url": "https://api.github.com/users/janekl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/janekl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/janekl/subscriptions",
"organizations_url": "https://api.github.com/users/janekl/orgs",
"repos_url": "https://api.github.com/users/janekl/repos",
"events_url": "https://api.github.com/users/janekl/events{/privacy}",
"received_events_url": "https://api.github.com/users/janekl/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7935/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/7935/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7934 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7934/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7934/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7934/events | https://github.com/NVIDIA/NeMo/pull/7934 | 2,008,777,718 | PR_kwDOC_bI7s5gQkna | 7,934 | add fim support | {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"The FIM part looks good to me. Thank you!",
"jenkins",
"jenkins",
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been ... | 2023-11-23T20:52:42 | 2024-02-02T22:23:39 | 2024-01-03T01:45:51 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7934",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7934",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7934.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7934.patch",
"merged_at": null
} | # What does this PR do ?
Adds FIM support for GPTDataset
Increases max number of training data files
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7934/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/7934/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7933 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7933/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7933/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7933/events | https://github.com/NVIDIA/NeMo/pull/7933 | 2,008,702,046 | PR_kwDOC_bI7s5gQUYV | 7,933 | Pass in rotary_base to mcore and from HF | {
"login": "Kipok",
"id": 2354422,
"node_id": "MDQ6VXNlcjIzNTQ0MjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/2354422?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Kipok",
"html_url": "https://github.com/Kipok",
"followers_url": "https://api.github.com/users/Kipok/followers",
"following_url": "https://api.github.com/users/Kipok/following{/other_user}",
"gists_url": "https://api.github.com/users/Kipok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Kipok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Kipok/subscriptions",
"organizations_url": "https://api.github.com/users/Kipok/orgs",
"repos_url": "https://api.github.com/users/Kipok/repos",
"events_url": "https://api.github.com/users/Kipok/events{/privacy}",
"received_events_url": "https://api.github.com/users/Kipok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"This PR incremented the version of Megatron Core from `e122536b7645edcb7ebf099b5c92a443f7dbf8e7` to `973330e9c3681604703bf1eb6b5a265d1b9b9b38`, that is from 0.3.0 to 0.4.0, but did not change the [requirements file](https://github.com/NVIDIA/NeMo/blob/main/requirements/requirements_nlp.tx... | 2023-11-23T19:10:37 | 2023-12-05T01:34:33 | 2023-12-03T23:00:42 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7933",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7933",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7933.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7933.patch",
"merged_at": "2023-12-03T23:00:42"
} | # What does this PR do ?
This depends on the following mcore commit https://github.com/NVIDIA/Megatron-LM/commit/d931ba8a492801d9d74efba6be94f82387090e57, so if it's not yet integrated into our main branch, we'd need to wait before merging this PR.
**Collection**: NLP
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7933/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/7933/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7932 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7932/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7932/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7932/events | https://github.com/NVIDIA/NeMo/issues/7932 | 2,008,530,188 | I_kwDOC_bI7s53t70M | 7,932 | How to continue finetuning of .nemo model from 768th epoch? | {
"login": "mukherjeesougata",
"id": 95704630,
"node_id": "U_kgDOBbRWNg",
"avatar_url": "https://avatars.githubusercontent.com/u/95704630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mukherjeesougata",
"html_url": "https://github.com/mukherjeesougata",
"followers_url": "https://api.github.com/users/mukherjeesougata/followers",
"following_url": "https://api.github.com/users/mukherjeesougata/following{/other_user}",
"gists_url": "https://api.github.com/users/mukherjeesougata/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mukherjeesougata/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mukherjeesougata/subscriptions",
"organizations_url": "https://api.github.com/users/mukherjeesougata/orgs",
"repos_url": "https://api.github.com/users/mukherjeesougata/repos",
"events_url": "https://api.github.com/users/mukherjeesougata/events{/privacy}",
"received_events_url": "https://api.github.com/users/mukherjeesougata/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | 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://a... | null | [
"@okuchaiev a gentle reminder for the same",
"If you have the two resume flags set, simply calling the script multiple times with the same paths will be sufficient ",
"Yes, I have tried that but it is restarting from the 0th epoch.",
"Please check the folder called nemo_experiments and if it has a folder insi... | 2023-11-23T16:22:21 | 2023-12-25T21:07:20 | 2023-12-25T21:07:19 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am using https://github.com/Open-Speech-EkStep/vakyansh-nemo-experimentation/blob/main/src/finetuning/speech_to_text_ctc_bpe.py to finetune. Due to power cut the process stopped after 768 epochs. So, I want to continue training the .nemo model from the 768th epoch. What extra arguments shall I pass
to speech_to_text_ctc_bpe.py for that?
My .sh file is as follows:-
```
if [[ ${tokenizer_type}=bpe ]]; then
CUDA_VISIBLE_DEVICES=3,4 python /home/conformer/vakyansh-nemo-experimentation/src/finetuning/speech_to_text_ctc_bpe.py \
--config-path=${config_path} \
--config-name=${config_name} \
model.train_ds.manifest_filepath=${train_manifest_path} \
model.validation_ds.manifest_filepath=${valid_manifest_path} \
model.test_ds.manifest_filepath=${test_manifest_path} \
model.tokenizer.dir=${tokenizer_dir} \
model.tokenizer.type=${tokenizer_type} \
trainer.devices=-1 \
trainer.accelerator="gpu" \
trainer.strategy="ddp" \
exp_manager.resume_if_exists=true \
exp_manager.resume_ignore_no_checkpoint=true \
exp_manager.create_wandb_logger=True \
exp_manager.wandb_logger_kwargs.name=${wandb_run_name} \
exp_manager.wandb_logger_kwargs.project=${wandb_project_name} \
+init_from_pretrained_model=${pretrained_model}
fi
```
```[tasklist]
### Tasks
```
| {
"login": "mukherjeesougata",
"id": 95704630,
"node_id": "U_kgDOBbRWNg",
"avatar_url": "https://avatars.githubusercontent.com/u/95704630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mukherjeesougata",
"html_url": "https://github.com/mukherjeesougata",
"followers_url": "https://api.github.com/users/mukherjeesougata/followers",
"following_url": "https://api.github.com/users/mukherjeesougata/following{/other_user}",
"gists_url": "https://api.github.com/users/mukherjeesougata/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mukherjeesougata/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mukherjeesougata/subscriptions",
"organizations_url": "https://api.github.com/users/mukherjeesougata/orgs",
"repos_url": "https://api.github.com/users/mukherjeesougata/repos",
"events_url": "https://api.github.com/users/mukherjeesougata/events{/privacy}",
"received_events_url": "https://api.github.com/users/mukherjeesougata/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7932/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/7932/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7931 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7931/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7931/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7931/events | https://github.com/NVIDIA/NeMo/pull/7931 | 2,007,424,111 | PR_kwDOC_bI7s5gMCxZ | 7,931 | Added YAML parsing for UB/TP overlap configs | {
"login": "sanandaraj5597",
"id": 55270934,
"node_id": "MDQ6VXNlcjU1MjcwOTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/55270934?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sanandaraj5597",
"html_url": "https://github.com/sanandaraj5597",
"followers_url": "https://api.github.com/users/sanandaraj5597/followers",
"following_url": "https://api.github.com/users/sanandaraj5597/following{/other_user}",
"gists_url": "https://api.github.com/users/sanandaraj5597/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sanandaraj5597/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sanandaraj5597/subscriptions",
"organizations_url": "https://api.github.com/users/sanandaraj5597/orgs",
"repos_url": "https://api.github.com/users/sanandaraj5597/repos",
"events_url": "https://api.github.com/users/sanandaraj5597/events{/privacy}",
"received_events_url": "https://api.github.com/users/sanandaraj5597/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"This PR aims to reverse changes from this PR: https://github.com/NVIDIA/NeMo/pull/7003\r\n\"Added a YAML parsing for UB/TP overlap configurations.\" doesn't describe the problem that you want to solve. Please be more verbose.\r\nThe https://github.com/NVIDIA/NeMo/pull/7003 removed module loading, open... | 2023-11-23T03:23:14 | 2023-12-02T02:45:50 | 2023-12-02T02:45:50 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7931",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7931",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7931.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7931.patch",
"merged_at": null
} | # What does this PR do ?
Added a YAML parsing for UB/TP overlap configurations. | {
"login": "sanandaraj5597",
"id": 55270934,
"node_id": "MDQ6VXNlcjU1MjcwOTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/55270934?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sanandaraj5597",
"html_url": "https://github.com/sanandaraj5597",
"followers_url": "https://api.github.com/users/sanandaraj5597/followers",
"following_url": "https://api.github.com/users/sanandaraj5597/following{/other_user}",
"gists_url": "https://api.github.com/users/sanandaraj5597/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sanandaraj5597/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sanandaraj5597/subscriptions",
"organizations_url": "https://api.github.com/users/sanandaraj5597/orgs",
"repos_url": "https://api.github.com/users/sanandaraj5597/repos",
"events_url": "https://api.github.com/users/sanandaraj5597/events{/privacy}",
"received_events_url": "https://api.github.com/users/sanandaraj5597/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7931/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/7931/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7929 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7929/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7929/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7929/events | https://github.com/NVIDIA/NeMo/issues/7929 | 2,005,472,112 | I_kwDOC_bI7s53iRNw | 7,929 | [BUG] Unittests rely on data from internal CI machines | {
"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
} | [
{
"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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | {
"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
} | [
{
"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:... | null | [
"Fix for this is #7943 \r\n\r\n",
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale.",
"```\r\nTraceback (most recent call last... | 2023-11-22T03:00:57 | 2024-05-28T15:10:23 | 2024-01-04T01:45:41 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
A lot of unit tests in NLP collection (over 10) require correct version of /home/TestData folder (from internal CI machines) to be present to run successfully.
This makes it impossible to run unittests successfully anywhere but on internal NVIDIA CI machines.
**To Reproduce**
Clone NeMo on new machine in clean environment and try running pytest tests/collections/nlp Make sure you do not have /home/TestData folder on the machine.
**Expected behavior**
Unittest run by pytest command should run successfully, not only on CI machines.
E.g. external developer/contributor should be able to run unit tests.
**Stack trace/logs**
```
library = 'sentencepiece', model_name = None, tokenizer_model = '/home/TestData/nlp/megatron_sft/tokenizer.model', vocab_file = None, merges_file = None, special_tokens = None, use_fast = False, bpe_dropout = 0.0, r2l = False, legacy = False, delimiter = None
def get_nmt_tokenizer(
library: str = 'yttm',
model_name: Optional[str] = None,
tokenizer_model: Optional[str] = None,
vocab_file: Optional[str] = None,
merges_file: Optional[str] = None,
special_tokens: Optional[Dict[str, str]] = None,
use_fast: Optional[bool] = False,
bpe_dropout: Optional[float] = 0.0,
r2l: Optional[bool] = False,
legacy: Optional[bool] = False,
delimiter: Optional[str] = None,
):
"""
Args:
model_name: if using a pretrained model from NeMo, HuggingFace, or Megatron
tokenizer_model: tokenizer model file of sentencepiece or youtokentome
special_tokens: dict of special tokens
vocab_file: path to vocab file
use_fast: (only for HuggingFace AutoTokenizer) set to True to use fast HuggingFace tokenizer
bpe_dropout: (only supported by YTTM tokenizer) BPE dropout tries to corrupt the standard segmentation procedure
of BPE to help model better learn word compositionality and become robust to segmentation errors.
It has empirically been shown to improve inference time BLEU scores.
r2l: Whether to return subword IDs from right to left
"""
if special_tokens is None:
special_tokens_dict = {}
else:
special_tokens_dict = special_tokens
if (library != 'byte-level') and (
model_name is None and (tokenizer_model is None or not os.path.isfile(tokenizer_model))
):
> raise ValueError("No Tokenizer path provided or file does not exist!")
E ValueError: No Tokenizer path provided or file does not exist!
nemo/collections/nlp/modules/common/tokenizer_utils.py:176: ValueError
```
**Environment (please complete the following information):**
PyTorch version 2.*
CUDA version
NCCL version
**Proposed fix**
I proposed tests to either skip if /home/TestData folder isn't found, or to be re-written.
Additional context
Add any other context about the problem here. | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7929/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/7929/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7928 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7928/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7928/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7928/events | https://github.com/NVIDIA/NeMo/pull/7928 | 2,005,458,711 | PR_kwDOC_bI7s5gFXqh | 7,928 | fix tp_overlap config var name | {
"login": "xrennvidia",
"id": 103958965,
"node_id": "U_kgDOBjJJtQ",
"avatar_url": "https://avatars.githubusercontent.com/u/103958965?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/xrennvidia",
"html_url": "https://github.com/xrennvidia",
"followers_url": "https://api.github.com/users/xrennvidia/followers",
"following_url": "https://api.github.com/users/xrennvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/xrennvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/xrennvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/xrennvidia/subscriptions",
"organizations_url": "https://api.github.com/users/xrennvidia/orgs",
"repos_url": "https://api.github.com/users/xrennvidia/repos",
"events_url": "https://api.github.com/users/xrennvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/xrennvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"LGTM. ",
"jenkins"
] | 2023-11-22T02:43:54 | 2023-11-22T07:19:09 | 2023-11-22T07:19:08 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7928",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7928",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7928.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7928.patch",
"merged_at": "2023-11-22T07:19:08"
} | # What does this PR do ?
TP overlap config var names in NeMo and MCore don't match. This PR will fix it.
| {
"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/7928/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/7928/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7927 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7927/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7927/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7927/events | https://github.com/NVIDIA/NeMo/pull/7927 | 2,005,399,459 | PR_kwDOC_bI7s5gFLMf | 7,927 | Fix for gpt3 eval hang with PP (a dtype issue) | {
"login": "yaoyu-33",
"id": 54727607,
"node_id": "MDQ6VXNlcjU0NzI3NjA3",
"avatar_url": "https://avatars.githubusercontent.com/u/54727607?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yaoyu-33",
"html_url": "https://github.com/yaoyu-33",
"followers_url": "https://api.github.com/users/yaoyu-33/followers",
"following_url": "https://api.github.com/users/yaoyu-33/following{/other_user}",
"gists_url": "https://api.github.com/users/yaoyu-33/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yaoyu-33/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yaoyu-33/subscriptions",
"organizations_url": "https://api.github.com/users/yaoyu-33/orgs",
"repos_url": "https://api.github.com/users/yaoyu-33/repos",
"events_url": "https://api.github.com/users/yaoyu-33/events{/privacy}",
"received_events_url": "https://api.github.com/users/yaoyu-33/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-22T01:24:54 | 2023-12-01T23:05:39 | 2023-12-01T23:05:38 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7927",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7927",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7927.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7927.patch",
"merged_at": "2023-12-01T23:05:38"
} | # What does this PR do ?
Fix for gpt3 eval hang with PP (a dtype issue)
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7927/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/7927/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7926 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7926/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7926/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7926/events | https://github.com/NVIDIA/NeMo/pull/7926 | 2,004,977,884 | PR_kwDOC_bI7s5gDvjw | 7,926 | Speedup RNN-T greedy decoding | {
"login": "artbataev",
"id": 9380560,
"node_id": "MDQ6VXNlcjkzODA1NjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/9380560?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/artbataev",
"html_url": "https://github.com/artbataev",
"followers_url": "https://api.github.com/users/artbataev/followers",
"following_url": "https://api.github.com/users/artbataev/following{/other_user}",
"gists_url": "https://api.github.com/users/artbataev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/artbataev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/artbataev/subscriptions",
"organizations_url": "https://api.github.com/users/artbataev/orgs",
"repos_url": "https://api.github.com/users/artbataev/repos",
"events_url": "https://api.github.com/users/artbataev/events{/privacy}",
"received_events_url": "https://api.github.com/users/artbataev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"... | 2023-11-21T19:18:55 | 2024-01-16T20:31:42 | 2024-01-16T20:31:41 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7926",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7926",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7926.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7926.patch",
"merged_at": "2024-01-16T20:31:41"
} | # What does this PR do ?
New algorithm for greedy batched decoding for RNN-Transducer.
With large batch sizes (e.g., 128) the expected speedup for large Fast Conformer-Transducer (full evaluation time including Encoder) is 1.7x-1.9x (when using `speech_to_text_eval.py`). For small batch sizes, e.g., 16, the observed speedup is ~1.3x.
The original algorithm is preserved and can be enabled by using `loop_labels=False`
E.g., on my local machine, with bf16, bs=128, Fast Conformer-Transducer Large, full `test-other` decoding
| Algorithm | Greedy | Greedy + Alignments |
| ---- | ---- | ---- |
| Current NeMo | 45 sec | 1 min 38 sec |
| Proposed | 24 sec | 30 sec |
**Collection**: [ASR]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
```shell
# default - new decoding algorithm
python examples/asr/speech_to_text_eval.py \
model_path=<nemo_model.nemo> \
dataset_manifest=<manifest> \
batch_size=128 \
output_filename=<output_mainfest_path>
# previous algorithm is preserved and can be used with `loop_labels=false`
python examples/asr/speech_to_text_eval.py \
model_path=<nemo_model.nemo> \
dataset_manifest=<manifest> \
batch_size=128 \
output_filename=<output_mainfest_path> \
rnnt_decoding.greedy.loop_labels=false
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "artbataev",
"id": 9380560,
"node_id": "MDQ6VXNlcjkzODA1NjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/9380560?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/artbataev",
"html_url": "https://github.com/artbataev",
"followers_url": "https://api.github.com/users/artbataev/followers",
"following_url": "https://api.github.com/users/artbataev/following{/other_user}",
"gists_url": "https://api.github.com/users/artbataev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/artbataev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/artbataev/subscriptions",
"organizations_url": "https://api.github.com/users/artbataev/orgs",
"repos_url": "https://api.github.com/users/artbataev/repos",
"events_url": "https://api.github.com/users/artbataev/events{/privacy}",
"received_events_url": "https://api.github.com/users/artbataev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7926/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/7926/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7925 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7925/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7925/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7925/events | https://github.com/NVIDIA/NeMo/pull/7925 | 2,004,875,262 | PR_kwDOC_bI7s5gDY7d | 7,925 | Fix pinned triton version | {
"login": "hsiehjackson",
"id": 37269846,
"node_id": "MDQ6VXNlcjM3MjY5ODQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/37269846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hsiehjackson",
"html_url": "https://github.com/hsiehjackson",
"followers_url": "https://api.github.com/users/hsiehjackson/followers",
"following_url": "https://api.github.com/users/hsiehjackson/following{/other_user}",
"gists_url": "https://api.github.com/users/hsiehjackson/gists{/gist_id}",
"starred_url": "https://api.github.com/users/hsiehjackson/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/hsiehjackson/subscriptions",
"organizations_url": "https://api.github.com/users/hsiehjackson/orgs",
"repos_url": "https://api.github.com/users/hsiehjackson/repos",
"events_url": "https://api.github.com/users/hsiehjackson/events{/privacy}",
"received_events_url": "https://api.github.com/users/hsiehjackson/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-21T18:13:33 | 2023-11-22T00:47:05 | 2023-11-22T00:47:04 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7925",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7925",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7925.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7925.patch",
"merged_at": "2023-11-22T00:47:04"
} | # What does this PR do ?
* Remove pinned version in Dockerfile.
* Raise error if triton version is incorrect when users use Flash Attention Triton.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7925/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/7925/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7924 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7924/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7924/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7924/events | https://github.com/NVIDIA/NeMo/pull/7924 | 2,003,913,499 | PR_kwDOC_bI7s5gAEtX | 7,924 | migration to PTL 2.0 for spellmapper model | {
"login": "bene-ges",
"id": 61418381,
"node_id": "MDQ6VXNlcjYxNDE4Mzgx",
"avatar_url": "https://avatars.githubusercontent.com/u/61418381?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bene-ges",
"html_url": "https://github.com/bene-ges",
"followers_url": "https://api.github.com/users/bene-ges/followers",
"following_url": "https://api.github.com/users/bene-ges/following{/other_user}",
"gists_url": "https://api.github.com/users/bene-ges/gists{/gist_id}",
"starred_url": "https://api.github.com/users/bene-ges/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/bene-ges/subscriptions",
"organizations_url": "https://api.github.com/users/bene-ges/orgs",
"repos_url": "https://api.github.com/users/bene-ges/repos",
"events_url": "https://api.github.com/users/bene-ges/events{/privacy}",
"received_events_url": "https://api.github.com/users/bene-ges/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"@athitten this change is similar to yours, could you review?",
"@titu1994, can you review or assign somebody?",
"@athitten can you review this?",
"jenkins",
"@ericharper, why my jenkins hangs forever? I don't have rights to see on it",
"jenkins",
"@titu1994, jenkins passed\r\n",
"@athitten, @ericharp... | 2023-11-21T10:08:54 | 2023-12-28T03:58:28 | 2023-12-28T03:58:27 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7924",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7924",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7924.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7924.patch",
"merged_at": "2023-12-28T03:58:27"
} | # What does this PR do ?
migration to PTL 2.0 for spellmapper model, same as in #7597
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "athitten",
"id": 47577437,
"node_id": "MDQ6VXNlcjQ3NTc3NDM3",
"avatar_url": "https://avatars.githubusercontent.com/u/47577437?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/athitten",
"html_url": "https://github.com/athitten",
"followers_url": "https://api.github.com/users/athitten/followers",
"following_url": "https://api.github.com/users/athitten/following{/other_user}",
"gists_url": "https://api.github.com/users/athitten/gists{/gist_id}",
"starred_url": "https://api.github.com/users/athitten/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/athitten/subscriptions",
"organizations_url": "https://api.github.com/users/athitten/orgs",
"repos_url": "https://api.github.com/users/athitten/repos",
"events_url": "https://api.github.com/users/athitten/events{/privacy}",
"received_events_url": "https://api.github.com/users/athitten/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7924/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/7924/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7923 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7923/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7923/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7923/events | https://github.com/NVIDIA/NeMo/pull/7923 | 2,003,504,509 | PR_kwDOC_bI7s5f-r1l | 7,923 | Add interface to set NCCL options of each process group | {
"login": "erhoo82",
"id": 15861259,
"node_id": "MDQ6VXNlcjE1ODYxMjU5",
"avatar_url": "https://avatars.githubusercontent.com/u/15861259?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/erhoo82",
"html_url": "https://github.com/erhoo82",
"followers_url": "https://api.github.com/users/erhoo82/followers",
"following_url": "https://api.github.com/users/erhoo82/following{/other_user}",
"gists_url": "https://api.github.com/users/erhoo82/gists{/gist_id}",
"starred_url": "https://api.github.com/users/erhoo82/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/erhoo82/subscriptions",
"organizations_url": "https://api.github.com/users/erhoo82/orgs",
"repos_url": "https://api.github.com/users/erhoo82/repos",
"events_url": "https://api.github.com/users/erhoo82/events{/privacy}",
"received_events_url": "https://api.github.com/users/erhoo82/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"@athitten Please review and merge"
] | 2023-11-21T05:41:47 | 2024-01-02T01:49:34 | 2023-12-04T19:23:08 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7923",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7923",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7923.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7923.patch",
"merged_at": "2023-12-04T19:23:08"
} | # What does this PR do ?
Add interface to set NCCL options of each process group. Can pass a path to yaml file that can specify the NCCL options (`min_ctas`, `max_ctas`, `cga_cluster_size`) for each process group.
Below shows an example yaml file that specifies config of data-parallel, pipeline-parallel, and tensor-parallel proc groups.
```
dp:
min_ctas: 4
max_ctas: 4
pp:
min_ctas: 2
max_ctas: 2
tp:
min_ctas: 24
max_ctas: 24
```
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "athitten",
"id": 47577437,
"node_id": "MDQ6VXNlcjQ3NTc3NDM3",
"avatar_url": "https://avatars.githubusercontent.com/u/47577437?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/athitten",
"html_url": "https://github.com/athitten",
"followers_url": "https://api.github.com/users/athitten/followers",
"following_url": "https://api.github.com/users/athitten/following{/other_user}",
"gists_url": "https://api.github.com/users/athitten/gists{/gist_id}",
"starred_url": "https://api.github.com/users/athitten/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/athitten/subscriptions",
"organizations_url": "https://api.github.com/users/athitten/orgs",
"repos_url": "https://api.github.com/users/athitten/repos",
"events_url": "https://api.github.com/users/athitten/events{/privacy}",
"received_events_url": "https://api.github.com/users/athitten/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7923/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/7923/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7922 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7922/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7922/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7922/events | https://github.com/NVIDIA/NeMo/pull/7922 | 2,003,127,981 | PR_kwDOC_bI7s5f9bI8 | 7,922 | fix issues with convert_nemo_llama_to_hf.py | {
"login": "Zhilin123",
"id": 29811458,
"node_id": "MDQ6VXNlcjI5ODExNDU4",
"avatar_url": "https://avatars.githubusercontent.com/u/29811458?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Zhilin123",
"html_url": "https://github.com/Zhilin123",
"followers_url": "https://api.github.com/users/Zhilin123/followers",
"following_url": "https://api.github.com/users/Zhilin123/following{/other_user}",
"gists_url": "https://api.github.com/users/Zhilin123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Zhilin123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Zhilin123/subscriptions",
"organizations_url": "https://api.github.com/users/Zhilin123/orgs",
"repos_url": "https://api.github.com/users/Zhilin123/repos",
"events_url": "https://api.github.com/users/Zhilin123/events{/privacy}",
"received_events_url": "https://api.github.com/users/Zhilin123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Do we have a test for `convert_nemo_llama_to_hf.py` anywhere?\r\n\r\nLike we have, for example, for the script `convert_hf_llama_to_nemo.py` working the other way around [here](https://github.com/NVIDIA/NeMo/blob/521cfb4c9359c4033f34776da4b73fabf3e94240/Jenkinsfile#L131-L138).",
"@janekl I'm not sure if there's ... | 2023-11-20T22:34:43 | 2023-11-25T20:27:25 | 2023-11-25T20:27:15 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7922",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7922",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7922.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7922.patch",
"merged_at": "2023-11-25T20:27:15"
} | # What does this PR do ?
fix issues with convert_nemo_llama_to_hf.py
**Collection**: nlp
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "Zhilin123",
"id": 29811458,
"node_id": "MDQ6VXNlcjI5ODExNDU4",
"avatar_url": "https://avatars.githubusercontent.com/u/29811458?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Zhilin123",
"html_url": "https://github.com/Zhilin123",
"followers_url": "https://api.github.com/users/Zhilin123/followers",
"following_url": "https://api.github.com/users/Zhilin123/following{/other_user}",
"gists_url": "https://api.github.com/users/Zhilin123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Zhilin123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Zhilin123/subscriptions",
"organizations_url": "https://api.github.com/users/Zhilin123/orgs",
"repos_url": "https://api.github.com/users/Zhilin123/repos",
"events_url": "https://api.github.com/users/Zhilin123/events{/privacy}",
"received_events_url": "https://api.github.com/users/Zhilin123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7922/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/7922/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7921 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7921/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7921/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7921/events | https://github.com/NVIDIA/NeMo/issues/7921 | 2,003,026,084 | I_kwDOC_bI7s53Y8Ck | 7,921 | Preprocessor cannot be exported to Core ML | {
"login": "msis",
"id": 577139,
"node_id": "MDQ6VXNlcjU3NzEzOQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/577139?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/msis",
"html_url": "https://github.com/msis",
"followers_url": "https://api.github.com/users/msis/followers",
"following_url": "https://api.github.com/users/msis/following{/other_user}",
"gists_url": "https://api.github.com/users/msis/gists{/gist_id}",
"starred_url": "https://api.github.com/users/msis/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/msis/subscriptions",
"organizations_url": "https://api.github.com/users/msis/orgs",
"repos_url": "https://api.github.com/users/msis/repos",
"events_url": "https://api.github.com/users/msis/events{/privacy}",
"received_events_url": "https://api.github.com/users/msis/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"A quick first win is to use `.size()` instead of `.shape[]` here https://github.com/NVIDIA/NeMo/blob/08937c8e7dd2e782fac99c6c230ae0170c277700/nemo/collections/asr/parts/preprocessing/features.py#L150 :\r\n\r\n```py\r\n mask = (\r\n torch.arange(like.size(time_dim), device=like.device)\r\n .repeat(... | 2023-11-20T21:21:14 | 2024-02-05T01:45:02 | 2024-02-05T01:45:01 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
The asr preprocessors cannot be exported to Core ML because of this function: `make_seq_mask_like`.
Any chance for a rewrite that allows the export?
**Steps/Code to reproduce bug**
Here I singled the function causing issues:
```python
import torch
import numpy as np
import coremltools as ct
from nemo.collections.asr.parts.preprocessing.features import make_seq_mask_like
_lengths = torch.load("lengths.pt")
_like = torch.load("like.pt")
_time_dim = torch.load("time_dim.pt")
_valid_ones = torch.load("valid_ones.pt")
_mask = torch.load("mask.pt")
mask = make_seq_mask_like(_lengths, _like, _time_dim, _valid_ones)
print(torch.all(torch.eq(mask, _mask)))
class SeqMaskLike(torch.nn.Module):
def __init__(self):
super().__init__()
def forward(self, lengths, like):
return make_seq_mask_like(lengths, like, _time_dim, _valid_ones)
model = SeqMaskLike()
model.eval()
traced_model = torch.jit.trace(model, [_lengths, _like])
# Convert to CoreML
model = ct.convert(
traced_model,
inputs=[
ct.TensorType(name="lengths", shape=_lengths.shape),
ct.TensorType(name="like", shape=_like.shape),
],
outputs=[ct.TensorType(name="mask")],
source="pytorch",
)
```
Here's the traceback of the error:
```
Traceback (most recent call last):
File "/Users/msis/Projects/tarteel/nemo2ios/poc/nemo_tracing_issue.py", line 30, in <module>
model = ct.convert(
^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/_converters_entry.py", line 574, in convert
mlmodel = mil_convert(
^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/converter.py", line 188, in mil_convert
return _mil_convert(model, convert_from, convert_to, ConverterRegistry, MLModel, compute_units, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/converter.py", line 212, in _mil_convert
proto, mil_program = mil_convert_to_proto(
^^^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/converter.py", line 286, in mil_convert_to_proto
prog = frontend_converter(model, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/converter.py", line 108, in __call__
return load(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 80, in load
return _perform_torch_convert(converter, debug)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 99, in _perform_torch_convert
prog = converter.convert()
^^^^^^^^^^^^^^^^^^^
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/frontend/torch/converter.py", line 519, in convert
convert_nodes(self.context, self.graph)
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 88, in convert_nodes
add_op(context, node)
File "/Users/msis/.virtualenvs/nemo2ios/lib/python3.11/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 3397, in getitem
raise AssertionError("Item selection is supported only on python list/tuple objects")
AssertionError: Item selection is supported only on python list/tuple objects
```
**Expected behavior**
Ideally, this should just work:
```python
import torch
from nemo.collections.asr.models import EncDecCTCModelBPE
import coremltools as ct
model = EncDecCTCModelBPE.from_pretrained(model_name="stt_en_conformer_ctc_large")
model.preprocessor.export("preprocessor.pt")
ts_model_preprocessor = torch.jit.load("preprocessor.pt")
input, input_length = model.preprocessor.input_example(max_batch=2) # 1 is the batch size
traced_processor = torch.jit.trace(model, [input, input_length])
ct_preprocessor = ct.convert(
traced_preprocessor,
inputs=[
ct.TensorType(name="input", shape=input.shape),
ct.TensorType(name="input_length", shape=input_length.shape),
],
)
```
**Environment overview (please complete the following information)**
- Environment location: macOS (venv)
- Method of NeMo install: `pip install nemo_toolkit[asr]`
**Environment details**
If NVIDIA docker image is used you don't need to specify these.
Otherwise, please provide:
- OS version: macOS 14.1.1
- PyTorch version: 2.1.1
- Python version: 3.11
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7921/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/7921/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7920 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7920/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7920/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7920/events | https://github.com/NVIDIA/NeMo/pull/7920 | 2,002,927,297 | PR_kwDOC_bI7s5f8u0u | 7,920 | Reworked MegatronPretrainingRandomBatchSampler to correctly handle epochs > 1 | {
"login": "trias702",
"id": 25867060,
"node_id": "MDQ6VXNlcjI1ODY3MDYw",
"avatar_url": "https://avatars.githubusercontent.com/u/25867060?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/trias702",
"html_url": "https://github.com/trias702",
"followers_url": "https://api.github.com/users/trias702/followers",
"following_url": "https://api.github.com/users/trias702/following{/other_user}",
"gists_url": "https://api.github.com/users/trias702/gists{/gist_id}",
"starred_url": "https://api.github.com/users/trias702/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/trias702/subscriptions",
"organizations_url": "https://api.github.com/users/trias702/orgs",
"repos_url": "https://api.github.com/users/trias702/repos",
"events_url": "https://api.github.com/users/trias702/events{/privacy}",
"received_events_url": "https://api.github.com/users/trias702/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-20T20:14:46 | 2023-12-10T01:41:53 | 2023-12-10T01:41:52 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7920",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7920",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7920.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7920.patch",
"merged_at": "2023-12-10T01:41:52"
} | # What does this PR do ?
Fixes MegatronPretrainingRandomBatchSampler to correctly instantiate when consumed_samples >= total_samples
**Collection**: NLP
# Changelog
- nemo/collections/nlp/data/language_modeling/megatron/megatron_batch_samplers.py: slightly rewrote the MegatronPretrainingRandomBatchSampler class
# Usage
No change to how it was used before, however, I have run many training runs using this fixed version in Nemo-Aligner and it appears to work correctly with no issues.
# Before your PR is "Ready for review"
**Pre checks**:
- [X] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [X] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7920/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/7920/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7918 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7918/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7918/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7918/events | https://github.com/NVIDIA/NeMo/issues/7918 | 2,002,454,648 | I_kwDOC_bI7s53Wwh4 | 7,918 | ModuleNotFoundError: No module named 'nemo.deploy' | {
"login": "deadsoul44",
"id": 31016182,
"node_id": "MDQ6VXNlcjMxMDE2MTgy",
"avatar_url": "https://avatars.githubusercontent.com/u/31016182?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/deadsoul44",
"html_url": "https://github.com/deadsoul44",
"followers_url": "https://api.github.com/users/deadsoul44/followers",
"following_url": "https://api.github.com/users/deadsoul44/following{/other_user}",
"gists_url": "https://api.github.com/users/deadsoul44/gists{/gist_id}",
"starred_url": "https://api.github.com/users/deadsoul44/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/deadsoul44/subscriptions",
"organizations_url": "https://api.github.com/users/deadsoul44/orgs",
"repos_url": "https://api.github.com/users/deadsoul44/repos",
"events_url": "https://api.github.com/users/deadsoul44/events{/privacy}",
"received_events_url": "https://api.github.com/users/deadsoul44/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | {
"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
} | [
{
"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:... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-20T15:27:35 | 2023-12-29T01:43:20 | 2023-12-29T01:43:20 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
I am trying to experiment with Nemotron model using the example in Huggingface.
https://huggingface.co/nvidia/nemotron-3-8b-qa-4k
I installed nemo but I get the following error.
ModuleNotFoundError Traceback (most recent call last)
Cell In[3], line 1
----> 1 from nemo.deploy import NemoQuery
2 # In this case, we run inference on the same machine
3 nq = NemoQuery(url="localhost:8000", model_name="Nemotron-3-8B-QA-4K")
ModuleNotFoundError: No module named 'nemo.deploy'
I am able to import nemo itself.
**Expected behavior**
Import nemo.deploy
**Environment overview (please complete the following information)**
- Environment location: Bare-metal
- Method of NeMo install: pip install in a conda env
- 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: MacOS
- PyTorch version: 2.1.1
- Python version: 3.10.12
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7918/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/7918/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7917 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7917/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7917/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7917/events | https://github.com/NVIDIA/NeMo/pull/7917 | 2,002,443,033 | PR_kwDOC_bI7s5f7FDH | 7,917 | Remove deprecated arguments from TE's TransformerLayer | {
"login": "jbaczek",
"id": 45043825,
"node_id": "MDQ6VXNlcjQ1MDQzODI1",
"avatar_url": "https://avatars.githubusercontent.com/u/45043825?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbaczek",
"html_url": "https://github.com/jbaczek",
"followers_url": "https://api.github.com/users/jbaczek/followers",
"following_url": "https://api.github.com/users/jbaczek/following{/other_user}",
"gists_url": "https://api.github.com/users/jbaczek/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbaczek/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbaczek/subscriptions",
"organizations_url": "https://api.github.com/users/jbaczek/orgs",
"repos_url": "https://api.github.com/users/jbaczek/repos",
"events_url": "https://api.github.com/users/jbaczek/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbaczek/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"the PR changes look fine but we might want to note that this only changes things for NeMo local GPT models, not m-core based ones which are the ones we majorly use now",
"I think the main issue is that user are used to configuring this from the config. Now TE expects an environment variable to be set. Mcore make... | 2023-11-20T15:21:57 | 2023-12-12T19:44:06 | 2023-12-12T19:44:06 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7917",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7917",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7917.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7917.patch",
"merged_at": "2023-12-12T19:44:06"
} | # What does this PR do ?
Transformer engine doesn't use this arguments since v0.7 and now in v1.0 they removed them completely.
**Collection**: [Note which collection this PR will affect]
NLP
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x ] Bugfix
- [ ] Documentation
## Who can review?
@ericharper
| {
"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/7917/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/7917/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7916 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7916/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7916/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7916/events | https://github.com/NVIDIA/NeMo/pull/7916 | 2,002,395,775 | PR_kwDOC_bI7s5f66s3 | 7,916 | [NLP] Access scaler only in FP16 case | {
"login": "janekl",
"id": 5582967,
"node_id": "MDQ6VXNlcjU1ODI5Njc=",
"avatar_url": "https://avatars.githubusercontent.com/u/5582967?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/janekl",
"html_url": "https://github.com/janekl",
"followers_url": "https://api.github.com/users/janekl/followers",
"following_url": "https://api.github.com/users/janekl/following{/other_user}",
"gists_url": "https://api.github.com/users/janekl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/janekl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/janekl/subscriptions",
"organizations_url": "https://api.github.com/users/janekl/orgs",
"repos_url": "https://api.github.com/users/janekl/repos",
"events_url": "https://api.github.com/users/janekl/events{/privacy}",
"received_events_url": "https://api.github.com/users/janekl/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-20T15:00:31 | 2023-12-07T08:21:11 | 2023-12-07T08:21:10 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7916",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7916",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7916.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7916.patch",
"merged_at": "2023-12-07T08:21:10"
} | # What does this PR do ?
When trying to load FP16 model with `trainer.precision != 16` (or `16-mixed`) a user gets a mysterious error message like:
```
File "/opt/NeMo/nemo/collections/nlp/models/language_modeling/megatron_base_model.py", line 122, in __init__
self.model_parallel_config: ModelParallelConfig = self.build_model_parallel_config()
File "/opt/NeMo/nemo/collections/nlp/models/language_modeling/megatron_base_model.py", line 782, in build_model_parallel_config
"grad_scale_func": self.trainer.precision_plugin.scaler.scale
AttributeError: 'NoneType' object has no attribute 'scale'
```
I'm changing the code to access `trainer.precision_plugin.scaler.scale` only when `trainer` is configured with the relevant precision: "16" or "16-mixed".
**Collection**: NLP
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [x] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [x] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas. | {
"login": "janekl",
"id": 5582967,
"node_id": "MDQ6VXNlcjU1ODI5Njc=",
"avatar_url": "https://avatars.githubusercontent.com/u/5582967?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/janekl",
"html_url": "https://github.com/janekl",
"followers_url": "https://api.github.com/users/janekl/followers",
"following_url": "https://api.github.com/users/janekl/following{/other_user}",
"gists_url": "https://api.github.com/users/janekl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/janekl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/janekl/subscriptions",
"organizations_url": "https://api.github.com/users/janekl/orgs",
"repos_url": "https://api.github.com/users/janekl/repos",
"events_url": "https://api.github.com/users/janekl/events{/privacy}",
"received_events_url": "https://api.github.com/users/janekl/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7916/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/7916/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7915 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7915/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7915/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7915/events | https://github.com/NVIDIA/NeMo/issues/7915 | 2,002,003,046 | I_kwDOC_bI7s53VCRm | 7,915 | How to convert distributed Megatron-LM ckpt to nemo | {
"login": "MachineGunLin",
"id": 18243223,
"node_id": "MDQ6VXNlcjE4MjQzMjIz",
"avatar_url": "https://avatars.githubusercontent.com/u/18243223?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MachineGunLin",
"html_url": "https://github.com/MachineGunLin",
"followers_url": "https://api.github.com/users/MachineGunLin/followers",
"following_url": "https://api.github.com/users/MachineGunLin/following{/other_user}",
"gists_url": "https://api.github.com/users/MachineGunLin/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MachineGunLin/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MachineGunLin/subscriptions",
"organizations_url": "https://api.github.com/users/MachineGunLin/orgs",
"repos_url": "https://api.github.com/users/MachineGunLin/repos",
"events_url": "https://api.github.com/users/MachineGunLin/events{/privacy}",
"received_events_url": "https://api.github.com/users/MachineGunLin/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 | [
"Hi @MachineGunLin \r\n\r\nCan you please let us know how you resolved the issue?"
] | 2023-11-20T11:40:27 | 2023-12-14T12:37:39 | 2023-11-21T03:26:55 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
I have tried to convert distributed Megatron-LM ckpt to NeMo, the checkpoint files are below (tensor-parallel-size=2, pipeline-model-parallel-size=2):
<img width="624" alt="image" src="https://github.com/NVIDIA/NeMo/assets/18243223/b8bb5696-5418-4a6d-8edc-3612ca7b812c">
I used this command to convert ckpt format:
```bash
(llama2) linrongjian@infiniAI-nxdx-B01-08:~/NeMo$ torchrun --nproc_per_node=4 \
./examples/nlp/language_modeling/megatron_lm_ckpt_to_nemo.py \
--checkpoint_folder /home/linrongjian/Megatron-LM/gpt2/checkpoint/iter_0000010 \
--checkpoint_name model_optim_rng.pt \
--nemo_file_path \
/home/linrongjian/NeMo/examples/nlp/language_modeling/gpt_distributed.nemo \
--model_type gpt \
--tensor_model_parallel_size 2 \
--pipeline_model_parallel_size 2 \
--gpus_per_node=1 \
--hparams_file /home/linrongjian/NeMo/hparams_distributed.yaml
```
And the output is:
```bash
[2023-11-20 11:34:40,349] torch.distributed.run: [WARNING]
[2023-11-20 11:34:40,349] torch.distributed.run: [WARNING] *****************************************
[2023-11-20 11:34:40,349] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
[2023-11-20 11:34:40,349] torch.distributed.run: [WARNING] *****************************************
[NeMo I 2023-11-20 11:34:49 nemo_logging:381] Initializing torch.distributed with local_rank: 0, rank: 0, world_size: 4
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has data parallel group: [0]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] All data parallel group ranks: [[0], [1], [2], [3]]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Ranks 0 has data parallel rank: 0
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has model parallel group: [0, 1, 2, 3]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] All model parallel group ranks: [[0, 1, 2, 3]]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has tensor model parallel group: [0, 1]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] All tensor model parallel group ranks: [[0, 1], [2, 3]]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has tensor model parallel rank: 0
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has pipeline model parallel group: [0, 2]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has embedding group: [0, 2]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] All pipeline model parallel group ranks: [[0, 2], [1, 3]]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has pipeline model parallel rank 0
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] All embedding group ranks: [[0, 2], [1, 3]]
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] Rank 0 has embedding rank: 0
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
[NeMo I 2023-11-20 11:34:51 nemo_logging:381] loading checkpoint /home/linrongjian/Megatron-LM/gpt2/checkpoint/iter_0000010/mp_rank_00_000/model_optim_rng.pt
converted 102.48M parameters
converted 101.44M parameters
[NeMo W 2023-11-20 11:34:51 nemo_logging:393] the checkpoint version is 3.0
converted 101.44M parameters
converted 102.48M parameters
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has data parallel group: [0]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] All data parallel group ranks: [[0], [1], [2], [3]]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Ranks 0 has data parallel rank: 0
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has model parallel group: [0, 1, 2, 3]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] All model parallel group ranks: [[0, 1, 2, 3]]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has tensor model parallel group: [0, 1]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] All tensor model parallel group ranks: [[0, 1], [2, 3]]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has tensor model parallel rank: 0
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has pipeline model parallel group: [0, 2]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has embedding group: [0, 2]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] All pipeline model parallel group ranks: [[0, 2], [1, 3]]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has pipeline model parallel rank 0
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] All embedding group ranks: [[0, 2], [1, 3]]
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Rank 0 has embedding rank: 0
[NeMo W 2023-11-20 11:34:52 nemo_logging:393] You tried to register an artifact under config key=tokenizer.vocab_file but an artifact for it has already been registered.
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Getting Megatron tokenizer for pretrained model name: megatron-gpt-345m, custom vocab file: /home/linrongjian/Megatron-LM/gpt2/gpt2-vocab.json, and merges file: /home/linrongjian/Megatron-LM/gpt2/gpt2-merges.txt
[NeMo I 2023-11-20 11:34:52 nemo_logging:381] Getting HuggingFace AutoTokenizer with pretrained_model_name: gpt2, vocab_file: /home/linrongjian/Megatron-LM/gpt2/gpt2-vocab.json, merges_files: /home/linrongjian/Megatron-LM/gpt2/gpt2-merges.txt, special_tokens_dict: {}, and use_fast: False
Using sep_token, but it is not set yet.
Using cls_token, but it is not set yet.
Using pad_token, but it is not set yet.
Using mask_token, but it is not set yet.
Using sep_token, but it is not set yet.
Using cls_token, but it is not set yet.
Using pad_token, but it is not set yet.
Using mask_token, but it is not set yet.
[NeMo I 2023-11-20 11:35:22 nemo_logging:381] Padded vocab_size: 50432, original vocab_size: 50257, dummy tokens: 175.
hidden_size: 1024, num_attention_heads: 16
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] The model: MegatronGPTModel() does not have field.name: fp8_wgrad in its cfg. Add this key to cfg or config_mapping to make to make it configurable.
Using sep_token, but it is not set yet.
Using cls_token, but it is not set yet.
Using pad_token, but it is not set yet.
Using mask_token, but it is not set yet.
hidden_size: 1024, num_attention_heads: 16
hidden_size: 1024, num_attention_heads: 16
Using sep_token, but it is not set yet.
Using cls_token, but it is not set yet.
Using pad_token, but it is not set yet.
Using mask_token, but it is not set yet.
hidden_size: 1024, num_attention_heads: 16
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] /home/linrongjian/anaconda3/envs/llama2/lib/python3.11/site-packages/apex/transformer/layers/layer_norm.py:93: UserWarning: `apex.contrib.layer_norm.FastLayerNorm` isn't available thus falling back to `apex.normalization.FusedLayerNorm`
warnings.warn("`apex.contrib.layer_norm.FastLayerNorm` isn't available thus falling back to `apex.normalization.FusedLayerNorm`")
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] /home/linrongjian/anaconda3/envs/llama2/lib/python3.11/site-packages/pytorch_lightning/core/saving.py:161: UserWarning: Found keys that are in the model state dict but not in the checkpoint: ['model.language_model.encoder.layers.0.input_layernorm.weight', 'model.language_model.encoder.layers.0.input_layernorm.bias', 'model.language_model.encoder.layers.0.post_attention_layernorm.weight', 'model.language_model.encoder.layers.0.post_attention_layernorm.bias', 'model.language_model.encoder.layers.1.input_layernorm.weight', 'model.language_model.encoder.layers.1.input_layernorm.bias', 'model.language_model.encoder.layers.1.post_attention_layernorm.weight', 'model.language_model.encoder.layers.1.post_attention_layernorm.bias', 'model.language_model.encoder.layers.2.input_layernorm.weight', 'model.language_model.encoder.layers.2.input_layernorm.bias', 'model.language_model.encoder.layers.2.post_attention_layernorm.weight', 'model.language_model.encoder.layers.2.post_attention_layernorm.bias', 'model.language_model.encoder.layers.3.input_layernorm.weight', 'model.language_model.encoder.layers.3.input_layernorm.bias', 'model.language_model.encoder.layers.3.post_attention_layernorm.weight', 'model.language_model.encoder.layers.3.post_attention_layernorm.bias', 'model.language_model.encoder.layers.4.input_layernorm.weight', 'model.language_model.encoder.layers.4.input_layernorm.bias', 'model.language_model.encoder.layers.4.post_attention_layernorm.weight', 'model.language_model.encoder.layers.4.post_attention_layernorm.bias', 'model.language_model.encoder.layers.5.input_layernorm.weight', 'model.language_model.encoder.layers.5.input_layernorm.bias', 'model.language_model.encoder.layers.5.post_attention_layernorm.weight', 'model.language_model.encoder.layers.5.post_attention_layernorm.bias', 'model.language_model.encoder.layers.6.input_layernorm.weight', 'model.language_model.encoder.layers.6.input_layernorm.bias', 'model.language_model.encoder.layers.6.post_attention_layernorm.weight', 'model.language_model.encoder.layers.6.post_attention_layernorm.bias', 'model.language_model.encoder.layers.7.input_layernorm.weight', 'model.language_model.encoder.layers.7.input_layernorm.bias', 'model.language_model.encoder.layers.7.post_attention_layernorm.weight', 'model.language_model.encoder.layers.7.post_attention_layernorm.bias', 'model.language_model.encoder.layers.8.input_layernorm.weight', 'model.language_model.encoder.layers.8.input_layernorm.bias', 'model.language_model.encoder.layers.8.post_attention_layernorm.weight', 'model.language_model.encoder.layers.8.post_attention_layernorm.bias', 'model.language_model.encoder.layers.9.input_layernorm.weight', 'model.language_model.encoder.layers.9.input_layernorm.bias', 'model.language_model.encoder.layers.9.post_attention_layernorm.weight', 'model.language_model.encoder.layers.9.post_attention_layernorm.bias', 'model.language_model.encoder.layers.10.input_layernorm.weight', 'model.language_model.encoder.layers.10.input_layernorm.bias', 'model.language_model.encoder.layers.10.post_attention_layernorm.weight', 'model.language_model.encoder.layers.10.post_attention_layernorm.bias', 'model.language_model.encoder.layers.11.input_layernorm.weight', 'model.language_model.encoder.layers.11.input_layernorm.bias', 'model.language_model.encoder.layers.11.post_attention_layernorm.weight', 'model.language_model.encoder.layers.11.post_attention_layernorm.bias']
rank_zero_warn(
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] /home/linrongjian/anaconda3/envs/llama2/lib/python3.11/site-packages/pytorch_lightning/core/saving.py:165: UserWarning: Found keys that are not in the model state dict but in the checkpoint: ['model.language_model.encoder.layers.0.input_norm.weight', 'model.language_model.encoder.layers.0.input_norm.bias', 'model.language_model.encoder.layers.0.post_attention_norm.weight', 'model.language_model.encoder.layers.0.post_attention_norm.bias', 'model.language_model.encoder.layers.1.input_norm.weight', 'model.language_model.encoder.layers.1.input_norm.bias', 'model.language_model.encoder.layers.1.post_attention_norm.weight', 'model.language_model.encoder.layers.1.post_attention_norm.bias', 'model.language_model.encoder.layers.2.input_norm.weight', 'model.language_model.encoder.layers.2.input_norm.bias', 'model.language_model.encoder.layers.2.post_attention_norm.weight', 'model.language_model.encoder.layers.2.post_attention_norm.bias', 'model.language_model.encoder.layers.3.input_norm.weight', 'model.language_model.encoder.layers.3.input_norm.bias', 'model.language_model.encoder.layers.3.post_attention_norm.weight', 'model.language_model.encoder.layers.3.post_attention_norm.bias', 'model.language_model.encoder.layers.4.input_norm.weight', 'model.language_model.encoder.layers.4.input_norm.bias', 'model.language_model.encoder.layers.4.post_attention_norm.weight', 'model.language_model.encoder.layers.4.post_attention_norm.bias', 'model.language_model.encoder.layers.5.input_norm.weight', 'model.language_model.encoder.layers.5.input_norm.bias', 'model.language_model.encoder.layers.5.post_attention_norm.weight', 'model.language_model.encoder.layers.5.post_attention_norm.bias', 'model.language_model.encoder.layers.6.input_norm.weight', 'model.language_model.encoder.layers.6.input_norm.bias', 'model.language_model.encoder.layers.6.post_attention_norm.weight', 'model.language_model.encoder.layers.6.post_attention_norm.bias', 'model.language_model.encoder.layers.7.input_norm.weight', 'model.language_model.encoder.layers.7.input_norm.bias', 'model.language_model.encoder.layers.7.post_attention_norm.weight', 'model.language_model.encoder.layers.7.post_attention_norm.bias', 'model.language_model.encoder.layers.8.input_norm.weight', 'model.language_model.encoder.layers.8.input_norm.bias', 'model.language_model.encoder.layers.8.post_attention_norm.weight', 'model.language_model.encoder.layers.8.post_attention_norm.bias', 'model.language_model.encoder.layers.9.input_norm.weight', 'model.language_model.encoder.layers.9.input_norm.bias', 'model.language_model.encoder.layers.9.post_attention_norm.weight', 'model.language_model.encoder.layers.9.post_attention_norm.bias', 'model.language_model.encoder.layers.10.input_norm.weight', 'model.language_model.encoder.layers.10.input_norm.bias', 'model.language_model.encoder.layers.10.post_attention_norm.weight', 'model.language_model.encoder.layers.10.post_attention_norm.bias', 'model.language_model.encoder.layers.11.input_norm.weight', 'model.language_model.encoder.layers.11.input_norm.bias', 'model.language_model.encoder.layers.11.post_attention_norm.weight', 'model.language_model.encoder.layers.11.post_attention_norm.bias']
rank_zero_warn(
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] You tried to register an artifact under config key=tokenizer.vocab_file but an artifact for it has already been registered.
[NeMo W 2023-11-20 11:35:22 nemo_logging:393] You tried to register an artifact under config key=tokenizer.merge_file but an artifact for it has already been registered.
[NeMo I 2023-11-20 11:35:25 nemo_logging:381] NeMo model saved to: /home/linrongjian/NeMo/examples/nlp/language_modeling/gpt_distributed.nemo
```
Seems like only one ckpt(mp_rank_00_000) got convertted, while the other three didn't.
**Expected behavior**
Should get all nemo ckpts.
**Environment overview (please complete the following information)**
- Method of NeMo install: from source (main branch)
**Additional context**
GPU: Nvidia A100 * 8
| {
"login": "MachineGunLin",
"id": 18243223,
"node_id": "MDQ6VXNlcjE4MjQzMjIz",
"avatar_url": "https://avatars.githubusercontent.com/u/18243223?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MachineGunLin",
"html_url": "https://github.com/MachineGunLin",
"followers_url": "https://api.github.com/users/MachineGunLin/followers",
"following_url": "https://api.github.com/users/MachineGunLin/following{/other_user}",
"gists_url": "https://api.github.com/users/MachineGunLin/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MachineGunLin/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MachineGunLin/subscriptions",
"organizations_url": "https://api.github.com/users/MachineGunLin/orgs",
"repos_url": "https://api.github.com/users/MachineGunLin/repos",
"events_url": "https://api.github.com/users/MachineGunLin/events{/privacy}",
"received_events_url": "https://api.github.com/users/MachineGunLin/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7915/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/7915/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7914 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7914/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7914/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7914/events | https://github.com/NVIDIA/NeMo/issues/7914 | 2,001,768,025 | I_kwDOC_bI7s53UI5Z | 7,914 | How to get the data_process code in NeMo doc for common crawl? | {
"login": "SefaZeng",
"id": 13918630,
"node_id": "MDQ6VXNlcjEzOTE4NjMw",
"avatar_url": "https://avatars.githubusercontent.com/u/13918630?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SefaZeng",
"html_url": "https://github.com/SefaZeng",
"followers_url": "https://api.github.com/users/SefaZeng/followers",
"following_url": "https://api.github.com/users/SefaZeng/following{/other_user}",
"gists_url": "https://api.github.com/users/SefaZeng/gists{/gist_id}",
"starred_url": "https://api.github.com/users/SefaZeng/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/SefaZeng/subscriptions",
"organizations_url": "https://api.github.com/users/SefaZeng/orgs",
"repos_url": "https://api.github.com/users/SefaZeng/repos",
"events_url": "https://api.github.com/users/SefaZeng/events{/privacy}",
"received_events_url": "https://api.github.com/users/SefaZeng/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | 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://a... | null | [
"@SefaZeng there is a issue with links, its not open sourced through NeMo. \r\nCurrently you can request for access to NeMo framework at https://developer.nvidia.com/nemo-framework and run scripts",
"@SefaZeng closing this issue, feel free to reopen if you are unable to run the scripts after downloading the nemo-... | 2023-11-20T09:39:51 | 2023-11-21T17:21:52 | 2023-11-21T06:09:26 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Is your feature request related to a problem? Please describe.**
I read the doc for NeMO about data processing [here,](https://docs.nvidia.com/nemo-framework/user-guide/latest/modelguide/pretrainingdatasets/textcleaningandlanguageseparation.html) and I want to use these scripts for data preprocessing. But I can't find a way to get them.
**Describe the solution you'd like**
A website or a github URL to download all these scripts.
| {
"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/7914/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/7914/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7913 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7913/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7913/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7913/events | https://github.com/NVIDIA/NeMo/pull/7913 | 2,001,437,983 | PR_kwDOC_bI7s5f3rOY | 7,913 | added missing torch import | {
"login": "Davood-M",
"id": 3964285,
"node_id": "MDQ6VXNlcjM5NjQyODU=",
"avatar_url": "https://avatars.githubusercontent.com/u/3964285?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Davood-M",
"html_url": "https://github.com/Davood-M",
"followers_url": "https://api.github.com/users/Davood-M/followers",
"following_url": "https://api.github.com/users/Davood-M/following{/other_user}",
"gists_url": "https://api.github.com/users/Davood-M/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Davood-M/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Davood-M/subscriptions",
"organizations_url": "https://api.github.com/users/Davood-M/orgs",
"repos_url": "https://api.github.com/users/Davood-M/repos",
"events_url": "https://api.github.com/users/Davood-M/events{/privacy}",
"received_events_url": "https://api.github.com/users/Davood-M/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-20T06:08:53 | 2023-11-20T17:10:58 | 2023-11-20T17:10:57 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7913",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7913",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7913.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7913.patch",
"merged_at": "2023-11-20T17:10:57"
} | # What does this PR do ?
Added torch import for this error:
```
File "/code/nemo/collections/nlp/modules/common/megatron/adapters/mcore_mixins.py", line 85, in get_query_key_value_tensors
(query, key, value) = torch.split(
NameError: name 'torch' is not defined
```
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7913/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/7913/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7912 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7912/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7912/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7912/events | https://github.com/NVIDIA/NeMo/issues/7912 | 2,001,240,388 | I_kwDOC_bI7s53SIFE | 7,912 | Long-form audio speaker diarization OOM in clustering | {
"login": "remenberl",
"id": 2282568,
"node_id": "MDQ6VXNlcjIyODI1Njg=",
"avatar_url": "https://avatars.githubusercontent.com/u/2282568?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/remenberl",
"html_url": "https://github.com/remenberl",
"followers_url": "https://api.github.com/users/remenberl/followers",
"following_url": "https://api.github.com/users/remenberl/following{/other_user}",
"gists_url": "https://api.github.com/users/remenberl/gists{/gist_id}",
"starred_url": "https://api.github.com/users/remenberl/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/remenberl/subscriptions",
"organizations_url": "https://api.github.com/users/remenberl/orgs",
"repos_url": "https://api.github.com/users/remenberl/repos",
"events_url": "https://api.github.com/users/remenberl/events{/privacy}",
"received_events_url": "https://api.github.com/users/remenberl/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | {
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.github.com/users/tango4j/followers",
"following_url": "https://api.github.com/users/tango4j/following{/other_user}",
"gists_url": "https://api.github.com/users/tango4j/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tango4j/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tango4j/subscriptions",
"organizations_url": "https://api.github.com/users/tango4j/orgs",
"repos_url": "https://api.github.com/users/tango4j/repos",
"events_url": "https://api.github.com/users/tango4j/events{/privacy}",
"received_events_url": "https://api.github.com/users/tango4j/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.git... | null | [
"Thanks for raising this issue, could you please attach full log here. ",
"Attached the log from running NeuralDiarizer. \r\n\r\n[nemo.log](https://github.com/NVIDIA/NeMo/files/13428499/nemo.log)\r\n",
"I'm also having this issue, hangs at this step, similar settings as op. Versions: Nemo 1.21.0, python 3.10\r\... | 2023-11-20T02:24:34 | 2024-01-21T01:49:47 | 2024-01-21T01:49:47 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi,
Thanks for the recent development of long-form audio speaker diarization in https://github.com/NVIDIA/NeMo/pull/7737. Recently I encounter a 4-hour-long audio and observe OOM on RAM (not VRAM).
**Steps/Code to reproduce bug**
It can be reproduced by using audio https://podwise-hh.s3.us-west-1.amazonaws.com/0ef8d4f5beb504fae9f12272a83db030ab29c92f4e033eb86bfefe3d1668c7cf.m4a
My telephonic config file is just the default, with the clustering/msdd part pasted below:
```
clustering:
parameters:
oracle_num_speakers: False
max_num_speakers: 8
enhanced_count_thres: 80
max_rp_threshold: 0.25
sparse_search_volume: 30
maj_vote_spk_count: False
chunk_cluster_count: 50
embeddings_per_chunk: 10000
msdd_model:
model_path: diar_msdd_telephonic
parameters:
use_speaker_model_from_ckpt: True
infer_batch_size: 25
sigmoid_threshold: [0.7]
seq_eval_mode: False
split_infer: True
diar_window_length: 50
overlap_infer_spk_limit: 5
```
**Expected behavior**
The job stops after my screen prints the last iteration of "Extracting embeddings for Diarization" and the program quickly consumes close to 64GB memory from 20GB- in previous steps. FYI, here are the last lines of prints before being killed.
```
[NeMo I 2023-11-19 20:54:29 clustering_diarizer:343] Extracting embeddings for Diarization
[NeMo I 2023-11-19 20:54:29 collections:445] Filtered duration for loading collection is 0.00 hours.
[NeMo I 2023-11-19 20:54:29 collections:446] Dataset loaded with 52949 items, total duration of 7.25 hours.
[NeMo I 2023-11-19 20:54:29 collections:448] # 52949 files loaded accounting to # 1 labels
```
**Environment overview (please complete the following information)**
- Environment location: Ubuntu 22.04 with 64GB RAM, 4090 GPU
- Method of NeMo install: `python -m pip install git+https://github.com/NVIDIA/NeMo.git@main#egg=nemo_toolkit[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: Ubuntu 22.04
- PyTorch version: '2.0.1+cu117'
- Python version: 3.10
**Additional context**
Add any other context about the problem here.
Example: GPU model
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7912/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/7912/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7911 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7911/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7911/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7911/events | https://github.com/NVIDIA/NeMo/issues/7911 | 2,000,899,249 | I_kwDOC_bI7s53Q0yx | 7,911 | v1.21.0 is incompatible with nemo2riva 2.13.1 | {
"login": "msis",
"id": 577139,
"node_id": "MDQ6VXNlcjU3NzEzOQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/577139?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/msis",
"html_url": "https://github.com/msis",
"followers_url": "https://api.github.com/users/msis/followers",
"following_url": "https://api.github.com/users/msis/following{/other_user}",
"gists_url": "https://api.github.com/users/msis/gists{/gist_id}",
"starred_url": "https://api.github.com/users/msis/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/msis/subscriptions",
"organizations_url": "https://api.github.com/users/msis/orgs",
"repos_url": "https://api.github.com/users/msis/repos",
"events_url": "https://api.github.com/users/msis/events{/privacy}",
"received_events_url": "https://api.github.com/users/msis/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"I think `test` should not be in `all`.\r\nFor now, a workaround is:\r\n\r\n```\r\npip install \r\n nemo_toolkit[core]==1.21.0 \\\r\n nemo_toolkit[common]==1.21.0 \\\r\n nemo_toolkit[asr]==1.21.0 \\\r\n nemo_toolkit[nlp]==1.21.0 \\\r\n nemo_toolkit[tts]==1.21.0 \\\r\n nemo_toolkit[slu]==1.21.0 \\\... | 2023-11-19T15:35:04 | 2024-02-28T01:44:07 | 2024-02-28T01:44:06 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
`nemo_toolkit[all]` has dependencies that conflict with `nemo2riva`:
```
29.39 The conflict is caused by:
29.39 nemo-toolkit[all] 1.21.0 depends on black==19.10b0; extra == "all"
29.39 nvidia-eff 0.6.5 depends on black==22.3.0
29.39 nemo-toolkit[all] 1.21.0 depends on isort<6.0.0 and >5.1.0; extra == "all"
29.39 nvidia-eff 0.5.7 depends on isort<5
```
Not sure why `black` and `isort` are listed as dependencies...
**Steps/Code to reproduce bug**
```sh
pip install nemo_toolkit[all]==1.21.0 nemo2riva==2.31.1
```
**Expected behavior**
Support nemo2riva
**Environment overview (please complete the following information)**
`nvcr.io/nvidia/riva/riva-speech:2.31.1-servicemaker`
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7911/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/7911/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7910 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7910/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7910/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7910/events | https://github.com/NVIDIA/NeMo/pull/7910 | 2,000,180,359 | PR_kwDOC_bI7s5fzpLs | 7,910 | Fix for megatron_amp_O2 + MegatronGPTSFTModel | {
"login": "vysarge",
"id": 10239441,
"node_id": "MDQ6VXNlcjEwMjM5NDQx",
"avatar_url": "https://avatars.githubusercontent.com/u/10239441?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vysarge",
"html_url": "https://github.com/vysarge",
"followers_url": "https://api.github.com/users/vysarge/followers",
"following_url": "https://api.github.com/users/vysarge/following{/other_user}",
"gists_url": "https://api.github.com/users/vysarge/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vysarge/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vysarge/subscriptions",
"organizations_url": "https://api.github.com/users/vysarge/orgs",
"repos_url": "https://api.github.com/users/vysarge/repos",
"events_url": "https://api.github.com/users/vysarge/events{/privacy}",
"received_events_url": "https://api.github.com/users/vysarge/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"Closing in favor of more-compatible change on the NLPAdapterModelMixin side in #7971."
] | 2023-11-18T03:01:46 | 2024-03-12T22:36:51 | 2023-12-05T01:21:17 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7910",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7910",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7910.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7910.patch",
"merged_at": null
} | Currently MegatronGPTSFTModel inherits sharded_state_dict from NLPAdapterModelMixin, but when restoring weights from the distributed checkpoint it's expected that MegatronGPTModel's version of sharded_state_dict has been run, resulting in:
```...
File "/root/llm/NeMo/examples/nlp/language_modeling/tuning/megatron_gpt_peft_tuning.py", line 63, in main
model = MegatronGPTSFTModel.restore_from(cfg.model.restore_from_path, model_cfg, trainer=trainer)
File "/root/llm/NeMo/nemo/collections/nlp/models/nlp_model.py", line 465, in restore_from
return super().restore_from(
File "/root/llm/NeMo/nemo/core/classes/modelPT.py", line 442, in restore_from
instance = cls._save_restore_connector.restore_from(
File "/root/llm/NeMo/nemo/collections/nlp/parts/nlp_overrides.py", line 743, in restore_from
checkpoint = dist_checkpointing.load(
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/serialization.py", line 72, in load
sharded_objects, sharded_state_dict = load_sharded_objects(sharded_state_dict, checkpoint_dir)
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/serialization.py", line 123, in load_sharded_objects
return dict_list_map_inplace(load_sharded_object, sharded_objects), sharded_state_dict
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/dict_utils.py", line 155, in dict_list_map_inplace
x[k] = dict_list_map_inplace(f, v)
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/dict_utils.py", line 155, in dict_list_map_inplace
x[k] = dict_list_map_inplace(f, v)
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/dict_utils.py", line 159, in dict_list_map_inplace
return f(x)
File "/opt/Megatron-LM/megatron/core/dist_checkpointing/serialization.py", line 120, in load_sharded_object
loaded_obj = torch.load(load_path)
File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 988, in load
with _open_file_like(f, 'rb') as opened_file:
File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 437, in _open_file_like
return _open_file(name_or_buffer, mode)
File "/usr/local/lib/python3.10/dist-packages/torch/serialization.py", line 418, in __init__
super().__init__(open(name, mode))
FileNotFoundError: [Errno 2] No such file or directory: '/pretrained_models/Llama-2-70b-bf16-mcore/model_weights/model.module.decoder.layers.self_attention.linear_proj._extra_state/shard_0_80.pt'
```
This code change overrides sharded_state_dict in MegatronGPTSFTModel to call that method instead. Changes may also be needed to reconcile MegatronT5SFTModel. | {
"login": "vysarge",
"id": 10239441,
"node_id": "MDQ6VXNlcjEwMjM5NDQx",
"avatar_url": "https://avatars.githubusercontent.com/u/10239441?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vysarge",
"html_url": "https://github.com/vysarge",
"followers_url": "https://api.github.com/users/vysarge/followers",
"following_url": "https://api.github.com/users/vysarge/following{/other_user}",
"gists_url": "https://api.github.com/users/vysarge/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vysarge/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vysarge/subscriptions",
"organizations_url": "https://api.github.com/users/vysarge/orgs",
"repos_url": "https://api.github.com/users/vysarge/repos",
"events_url": "https://api.github.com/users/vysarge/events{/privacy}",
"received_events_url": "https://api.github.com/users/vysarge/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7910/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/7910/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7909 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7909/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7909/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7909/events | https://github.com/NVIDIA/NeMo/pull/7909 | 2,000,085,797 | PR_kwDOC_bI7s5fzVHr | 7,909 | Add distopt support for FP8 params and BF16 optimizer state | {
"login": "timmoon10",
"id": 4406448,
"node_id": "MDQ6VXNlcjQ0MDY0NDg=",
"avatar_url": "https://avatars.githubusercontent.com/u/4406448?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/timmoon10",
"html_url": "https://github.com/timmoon10",
"followers_url": "https://api.github.com/users/timmoon10/followers",
"following_url": "https://api.github.com/users/timmoon10/following{/other_user}",
"gists_url": "https://api.github.com/users/timmoon10/gists{/gist_id}",
"starred_url": "https://api.github.com/users/timmoon10/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/timmoon10/subscriptions",
"organizations_url": "https://api.github.com/users/timmoon10/orgs",
"repos_url": "https://api.github.com/users/timmoon10/repos",
"events_url": "https://api.github.com/users/timmoon10/events{/privacy}",
"received_events_url": "https://api.github.com/users/timmoon10/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
... | closed | false | null | [] | null | [
"I trained the llama model with 2 nodes using model.fp8=True, model.fp8_params=True, model.optim.name=distributed_fused_adam, model.optim.dtype=bf16.\r\nI got this error when saving checkpoint\r\n\r\n<img width=\"1421\" alt=\"Screenshot 2023-11-20 at 17 27 23\" src=\"https://github.com/NVIDIA/NeMo/assets/25007718/5... | 2023-11-17T23:39:03 | 2024-01-12T07:08:49 | 2024-01-12T07:08:48 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7909",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7909",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7909.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7909.patch",
"merged_at": "2024-01-12T07:08:48"
} | # What does this PR do ?
Adds support in the Apex distributed Adam optimizer to support FP8 parameters (using experimental FP8 tensors from Transformer Engine) and BF16 optimizer state.
**Collection**: NLP
# Changelog
- Adds distributed optimizer support for FP8 parameters
- Adds the option to initialize GPT with FP8 parameters
- Add support for non-FP32 distributed optimizer state
# Usage
Run GPT, e.g. with the config at https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/language_modeling/conf/megatron_gpt_config.yaml.
Enable FP8 support with `model.fp8=True`, FP8 parameters with `model.fp8_params=True`, the distributed optimizer with `model.optim.name=distributed_fused_adam`, and BF16 optimizer state with `model.optim.dtype=bf16`.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [x] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Equivalent to https://github.com/NVIDIA/NeMo/pull/7885 and https://github.com/NVIDIA/NeMo/pull/7895.
* Depends on https://github.com/NVIDIA/apex/pull/1749
* Depends on ~https://github.com/NVIDIA/TransformerEngine/pull/524~ https://github.com/NVIDIA/TransformerEngine/pull/529
* Uses Transformer Engine `Float8Tensor` added in https://github.com/NVIDIA/TransformerEngine/pull/452
* Closes https://github.com/NVIDIA/NeMo/pull/7469, closes https://github.com/NVIDIA/NeMo/pull/7565, closes https://github.com/NVIDIA/NeMo/pull/7885, closes https://github.com/NVIDIA/NeMo/pull/7895.
| {
"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/7909/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/7909/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7908 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7908/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7908/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7908/events | https://github.com/NVIDIA/NeMo/pull/7908 | 1,999,871,426 | PR_kwDOC_bI7s5fyl9E | 7,908 | upgrade to latest mcore and TE | {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-17T20:25:08 | 2023-11-17T23:01:24 | 2023-11-17T23:01:23 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7908",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7908",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7908.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7908.patch",
"merged_at": "2023-11-17T23:01:23"
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7908/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/7908/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7907 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7907/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7907/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7907/events | https://github.com/NVIDIA/NeMo/issues/7907 | 1,999,749,823 | I_kwDOC_bI7s53McK_ | 7,907 | Tools to parallelize Megatron tokenization | {
"login": "adammoody",
"id": 1105742,
"node_id": "MDQ6VXNlcjExMDU3NDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1105742?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/adammoody",
"html_url": "https://github.com/adammoody",
"followers_url": "https://api.github.com/users/adammoody/followers",
"following_url": "https://api.github.com/users/adammoody/following{/other_user}",
"gists_url": "https://api.github.com/users/adammoody/gists{/gist_id}",
"starred_url": "https://api.github.com/users/adammoody/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/adammoody/subscriptions",
"organizations_url": "https://api.github.com/users/adammoody/orgs",
"repos_url": "https://api.github.com/users/adammoody/repos",
"events_url": "https://api.github.com/users/adammoody/events{/privacy}",
"received_events_url": "https://api.github.com/users/adammoody/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | closed | false | {
"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
} | [
{
"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:... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-17T19:06:34 | 2023-12-26T01:44:58 | 2023-12-26T01:44:57 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Is your feature request related to a problem? Please describe.**
I recently heard about the parallel data curation tools in NeMo, which allows one to run download, clean, and de-dup data using multiple nodes.
https://developer.nvidia.com/blog/curating-trillion-token-datasets-introducing-nemo-data-curator/
That's very exciting. Thanks for putting this together.
The NeMo script to tokenize data for Megatron supports running multiple workers on a single node:
https://github.com/NVIDIA/NeMo/blob/main/scripts/nlp_language_modeling/preprocess_data_for_megatron.py
For large datasets, I'd like to go even further and parallelize this task across multiple nodes.
**Describe the solution you'd like**
About a year ago, I wrote up some tools to do this. I opened a similar ticket on Megatron-LM with more info:
https://github.com/NVIDIA/Megatron-LM/issues/492
Would you be willing to accept something like this in NeMo?
Those existing files would need to be refreshed, because the indexed_dataset mmap file format has since changed. I am willing to take some time to do that if that would be helpful. | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7907/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/7907/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7906 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7906/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7906/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7906/events | https://github.com/NVIDIA/NeMo/pull/7906 | 1,998,574,407 | PR_kwDOC_bI7s5fuKy6 | 7,906 | add_support_for_swiglu_fusion | {
"login": "lhb8125",
"id": 16619280,
"node_id": "MDQ6VXNlcjE2NjE5Mjgw",
"avatar_url": "https://avatars.githubusercontent.com/u/16619280?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lhb8125",
"html_url": "https://github.com/lhb8125",
"followers_url": "https://api.github.com/users/lhb8125/followers",
"following_url": "https://api.github.com/users/lhb8125/following{/other_user}",
"gists_url": "https://api.github.com/users/lhb8125/gists{/gist_id}",
"starred_url": "https://api.github.com/users/lhb8125/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/lhb8125/subscriptions",
"organizations_url": "https://api.github.com/users/lhb8125/orgs",
"repos_url": "https://api.github.com/users/lhb8125/repos",
"events_url": "https://api.github.com/users/lhb8125/events{/privacy}",
"received_events_url": "https://api.github.com/users/lhb8125/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-17T08:50:02 | 2023-12-09T01:45:13 | 2023-12-09T01:45:13 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7906",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7906",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7906.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7906.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7906/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/7906/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7905 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7905/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7905/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7905/events | https://github.com/NVIDIA/NeMo/pull/7905 | 1,998,296,169 | PR_kwDOC_bI7s5ftPAN | 7,905 | Upgrade to PTL 2.1 | {
"login": "athitten",
"id": 47577437,
"node_id": "MDQ6VXNlcjQ3NTc3NDM3",
"avatar_url": "https://avatars.githubusercontent.com/u/47577437?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/athitten",
"html_url": "https://github.com/athitten",
"followers_url": "https://api.github.com/users/athitten/followers",
"following_url": "https://api.github.com/users/athitten/following{/other_user}",
"gists_url": "https://api.github.com/users/athitten/gists{/gist_id}",
"starred_url": "https://api.github.com/users/athitten/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/athitten/subscriptions",
"organizations_url": "https://api.github.com/users/athitten/orgs",
"repos_url": "https://api.github.com/users/athitten/repos",
"events_url": "https://api.github.com/users/athitten/events{/privacy}",
"received_events_url": "https://api.github.com/users/athitten/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale... | 2023-11-17T05:16:32 | 2023-12-13T01:46:03 | 2023-12-13T01:46:03 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7905",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7905",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7905.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7905.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7905/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/7905/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7904 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7904/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7904/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7904/events | https://github.com/NVIDIA/NeMo/pull/7904 | 1,998,093,643 | PR_kwDOC_bI7s5fsj7u | 7,904 | Pad sequences to multiples of 16 for GPTSFTDataset | {
"login": "vysarge",
"id": 10239441,
"node_id": "MDQ6VXNlcjEwMjM5NDQx",
"avatar_url": "https://avatars.githubusercontent.com/u/10239441?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vysarge",
"html_url": "https://github.com/vysarge",
"followers_url": "https://api.github.com/users/vysarge/followers",
"following_url": "https://api.github.com/users/vysarge/following{/other_user}",
"gists_url": "https://api.github.com/users/vysarge/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vysarge/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vysarge/subscriptions",
"organizations_url": "https://api.github.com/users/vysarge/orgs",
"repos_url": "https://api.github.com/users/vysarge/repos",
"events_url": "https://api.github.com/users/vysarge/events{/privacy}",
"received_events_url": "https://api.github.com/users/vysarge/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-17T01:36:56 | 2023-11-18T03:29:57 | 2023-11-17T20:10:05 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7904",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7904",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7904.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7904.patch",
"merged_at": "2023-11-17T20:10:05"
} | One-line change to pad sequences to the next highest multiple of 16 instead of 8 in GPTSFTDataset to allow this dataset's outputs to work with TransformerEngine which expects multiples of 16 when using fp8. | {
"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/7904/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/7904/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7903 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7903/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7903/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7903/events | https://github.com/NVIDIA/NeMo/pull/7903 | 1,997,974,270 | PR_kwDOC_bI7s5fsK_G | 7,903 | spelling mistake | {
"login": "orena1",
"id": 8983713,
"node_id": "MDQ6VXNlcjg5ODM3MTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/8983713?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/orena1",
"html_url": "https://github.com/orena1",
"followers_url": "https://api.github.com/users/orena1/followers",
"following_url": "https://api.github.com/users/orena1/following{/other_user}",
"gists_url": "https://api.github.com/users/orena1/gists{/gist_id}",
"starred_url": "https://api.github.com/users/orena1/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/orena1/subscriptions",
"organizations_url": "https://api.github.com/users/orena1/orgs",
"repos_url": "https://api.github.com/users/orena1/repos",
"events_url": "https://api.github.com/users/orena1/events{/privacy}",
"received_events_url": "https://api.github.com/users/orena1/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"Thanks !"
] | 2023-11-16T23:33:31 | 2023-12-24T22:56:52 | 2023-12-12T15:57:32 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7903",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7903",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7903.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7903.patch",
"merged_at": "2023-12-12T15:57:32"
} | # What does this PR do ?
spelling mistake
| {
"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/7903/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/7903/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7902 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7902/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7902/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7902/events | https://github.com/NVIDIA/NeMo/pull/7902 | 1,997,878,202 | PR_kwDOC_bI7s5fr1zo | 7,902 | Added knob for ub_tp_comm_overlap for the MCORE pass | {
"login": "sanandaraj5597",
"id": 55270934,
"node_id": "MDQ6VXNlcjU1MjcwOTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/55270934?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sanandaraj5597",
"html_url": "https://github.com/sanandaraj5597",
"followers_url": "https://api.github.com/users/sanandaraj5597/followers",
"following_url": "https://api.github.com/users/sanandaraj5597/following{/other_user}",
"gists_url": "https://api.github.com/users/sanandaraj5597/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sanandaraj5597/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sanandaraj5597/subscriptions",
"organizations_url": "https://api.github.com/users/sanandaraj5597/orgs",
"repos_url": "https://api.github.com/users/sanandaraj5597/repos",
"events_url": "https://api.github.com/users/sanandaraj5597/events{/privacy}",
"received_events_url": "https://api.github.com/users/sanandaraj5597/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-16T22:10:53 | 2023-11-17T02:57:01 | 2023-11-17T02:57:01 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7902",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7902",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7902.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7902.patch",
"merged_at": "2023-11-17T02:57:01"
} | The MCORE pass has been enabled to support User Buffer/Tensor Parallel communication overlap. That needs a knob at NeMo level which has been added. | {
"login": "sanandaraj5597",
"id": 55270934,
"node_id": "MDQ6VXNlcjU1MjcwOTM0",
"avatar_url": "https://avatars.githubusercontent.com/u/55270934?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sanandaraj5597",
"html_url": "https://github.com/sanandaraj5597",
"followers_url": "https://api.github.com/users/sanandaraj5597/followers",
"following_url": "https://api.github.com/users/sanandaraj5597/following{/other_user}",
"gists_url": "https://api.github.com/users/sanandaraj5597/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sanandaraj5597/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sanandaraj5597/subscriptions",
"organizations_url": "https://api.github.com/users/sanandaraj5597/orgs",
"repos_url": "https://api.github.com/users/sanandaraj5597/repos",
"events_url": "https://api.github.com/users/sanandaraj5597/events{/privacy}",
"received_events_url": "https://api.github.com/users/sanandaraj5597/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7902/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/7902/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7901 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7901/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7901/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7901/events | https://github.com/NVIDIA/NeMo/issues/7901 | 1,997,753,146 | I_kwDOC_bI7s53E0s6 | 7,901 | Chat method for TextGeneration API | {
"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
} | [
{
"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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8... | 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://a... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed becau... | 2023-11-16T20:48:39 | 2024-01-24T01:47:12 | 2024-01-24T01:47:11 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I propose ``class TextGeneration`` (from which ``MegatronGPTModel`` inherits) to add ```.chat(dict)``` method.
In addition to that **``MegatronGPTModel`` should have member method or properties which would allow to set or get model's template.** That template should be serialized/retrieved from ``model._cfg`` object and saved to .yaml file during model serialization.
Context. Base models do not require this method. But aligned models require setting the correct template for how user, assistant turns as well as system prompt need to be presented to the model. This means that the user of aligned .nemo checkpoint need to chase model's developer or documentation to understand what template they should set.
Example usage:
```python
messages = [
{"role": "system", "content": "You are a helpful AI assistant well trained in astrophysics. You produce helpful and insightful responses."},
{"role": "user", "content": "Does Earth revolve around the Sun?"},
{"role": "assistant", "content": "Yes, but to be more precise, Earth-Sun system revolves around a common center of mass which is few miles off Sun's center."},
{"role": "user", "content": "How many miles is it away from Sun's center?"}
]
model.eval()
responses=model.chat(messages)
print(responses)
```
If user calls ```.chat``` method on base model it should raise an exception saying chat template is not set.
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7901/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/7901/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7900 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7900/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7900/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7900/events | https://github.com/NVIDIA/NeMo/pull/7900 | 1,997,524,065 | PR_kwDOC_bI7s5fqm37 | 7,900 | Update manifest.py to speedup loading tarred datasets | {
"login": "stevehuang52",
"id": 105218074,
"node_id": "U_kgDOBkWAGg",
"avatar_url": "https://avatars.githubusercontent.com/u/105218074?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevehuang52",
"html_url": "https://github.com/stevehuang52",
"followers_url": "https://api.github.com/users/stevehuang52/followers",
"following_url": "https://api.github.com/users/stevehuang52/following{/other_user}",
"gists_url": "https://api.github.com/users/stevehuang52/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevehuang52/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevehuang52/subscriptions",
"organizations_url": "https://api.github.com/users/stevehuang52/orgs",
"repos_url": "https://api.github.com/users/stevehuang52/repos",
"events_url": "https://api.github.com/users/stevehuang52/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevehuang52/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4847373924,
"node_id": "LA_kwDOC_bI7s8AAAABIO0OZA",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/common",
"name": "common",
"color": "c5def5",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"> Is there any update on this ? @stevehuang52 @anteju\r\n\r\n@titu1994 Nope, I think it's ready to merge if there're no other comments.",
"jenkins"
] | 2023-11-16T18:48:29 | 2023-12-14T14:27:54 | 2023-12-14T14:27:53 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7900",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7900",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7900.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7900.patch",
"merged_at": "2023-12-14T14:27:53"
} | A simple trick to skip checking for relative filepaths when using tarred datasets. On my local desktop, the modification reduces the manifest loading time for tarred LS data from 4.98s to 3.28s, with 34% relative speedup. The trick is ugly...so please let me know of better ways to do it... | {
"login": "stevehuang52",
"id": 105218074,
"node_id": "U_kgDOBkWAGg",
"avatar_url": "https://avatars.githubusercontent.com/u/105218074?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stevehuang52",
"html_url": "https://github.com/stevehuang52",
"followers_url": "https://api.github.com/users/stevehuang52/followers",
"following_url": "https://api.github.com/users/stevehuang52/following{/other_user}",
"gists_url": "https://api.github.com/users/stevehuang52/gists{/gist_id}",
"starred_url": "https://api.github.com/users/stevehuang52/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/stevehuang52/subscriptions",
"organizations_url": "https://api.github.com/users/stevehuang52/orgs",
"repos_url": "https://api.github.com/users/stevehuang52/repos",
"events_url": "https://api.github.com/users/stevehuang52/events{/privacy}",
"received_events_url": "https://api.github.com/users/stevehuang52/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7900/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/7900/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7899 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7899/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7899/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7899/events | https://github.com/NVIDIA/NeMo/issues/7899 | 1,997,023,363 | I_kwDOC_bI7s53CCiD | 7,899 | I followed NVIDIA docs hub, and I got an error, `Exception: The provided qkv memory layout is not supported!` | {
"login": "kig1929",
"id": 11495946,
"node_id": "MDQ6VXNlcjExNDk1OTQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/11495946?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kig1929",
"html_url": "https://github.com/kig1929",
"followers_url": "https://api.github.com/users/kig1929/followers",
"following_url": "https://api.github.com/users/kig1929/following{/other_user}",
"gists_url": "https://api.github.com/users/kig1929/gists{/gist_id}",
"starred_url": "https://api.github.com/users/kig1929/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/kig1929/subscriptions",
"organizations_url": "https://api.github.com/users/kig1929/orgs",
"repos_url": "https://api.github.com/users/kig1929/repos",
"events_url": "https://api.github.com/users/kig1929/events{/privacy}",
"received_events_url": "https://api.github.com/users/kig1929/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-16T14:42:25 | 2023-12-25T01:45:45 | 2023-12-25T01:45:45 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
I'm trying to train with the nemo model following this [guide](https://docs.nvidia.com/nemo-framework/user-guide/latest/playbooks/llama2sft.html).
I've converted llama2-7b-hf to nemo model.
When I execute training script, I got this message `Exception: The provided qkv memory layout is not supported!`
**Steps/Code to reproduce bug**
I followed the guide from the beginning.
In the guide, followed [Step 4: Run the SFT finetuning script](https://docs.nvidia.com/nemo-framework/user-guide/latest/playbooks/llama2sft.html#step-4-run-the-sft-finetuning-script), only changing `megatron_amp_O2` from `True` to `False`.
(Without changing, I got FileNotFoundError regarding `.../model_weights/model.module. ... / shard_0_32.pt`)
And I got an error as below.
```
File "/home/projects/NeMo/examples/nlp/language_modeling/tuning/megatron_gpt_sft.py", line 226, in main
trainer.fit(model)
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py", line 532, in fit
call._call_and_handle_interrupt(
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/call.py", line 42, in _call_and_handle_interrupt
return trainer.strategy.launcher.launch(trainer_fn, *args, trainer=trainer, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/strategies/launchers/subprocess_script.py", line 93, in launch
return function(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py", line 571, in _fit_impl
self._run(model, ckpt_path=ckpt_path)
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py", line 980, in _run
results = self._run_stage()
^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py", line 1021, in _run_stage
self._run_sanity_check()
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/trainer.py", line 1050, in _run_sanity_check
val_loop.run()
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/loops/utilities.py", line 181, in _decorator
return loop_run(self, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/loops/evaluation_loop.py", line 115, in run
self._evaluation_step(batch, batch_idx, dataloader_idx)
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/loops/evaluation_loop.py", line 376, in _evaluation_step
output = call._call_strategy_hook(trainer, hook_name, *step_kwargs.values())
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/trainer/call.py", line 293, in _call_strategy_hook
output = fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/strategies/ddp.py", line 338, in validation_step
return self.model(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/torch/nn/parallel/distributed.py", line 1519, in forward
else self._run_ddp_forward(*inputs, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/torch/nn/parallel/distributed.py", line 1355, in _run_ddp_forward
return self.module(*inputs, **kwargs) # type: ignore[index]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/pytorch_lightning/overrides/base.py", line 102, in forward
return self._forward_module.validation_step(*inputs, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/projects/NeMo/nemo/collections/nlp/models/language_modeling/megatron_gpt_sft_model.py", line 387, in validation_step
return self.inference_step(dataloader_iter, batch_idx, 'validation', dataloader_idx)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/projects/NeMo/nemo/collections/nlp/models/language_modeling/megatron_gpt_sft_model.py", line 403, in inference_step
loss = super().validation_step(itertools.chain([batch]), batch_idx)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/projects/NeMo/nemo/collections/nlp/models/language_modeling/megatron_gpt_model.py", line 960, in validation_step
loss = self.fwd_bwd_step(dataloader_iter, batch_idx, True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/projects/NeMo/nemo/collections/nlp/models/language_modeling/megatron_gpt_sft_model.py", line 347, in fwd_bwd_step
losses_reduced_per_micro_batch = fwd_bwd_function(
^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/pipeline_parallel/schedules.py", line 329, in forward_backward_no_pipelining
output_tensor = forward_step(
^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/pipeline_parallel/schedules.py", line 185, in forward_step
output_tensor, loss_func = forward_step_func(data_iterator, model)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/projects/NeMo/nemo/collections/nlp/models/language_modeling/megatron_gpt_model.py", line 860, in fwd_output_and_loss_func
output_tensor = model(**forward_args)
^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/models/gpt/gpt_model.py", line 172, in forward
hidden_states = self.decoder(
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/transformer/transformer_block.py", line 250, in forward
hidden_states = layer(
^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/transformer/transformer_layer.py", line 124, in forward
attention_output_with_bias = self.self_attention(
^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/transformer/attention.py", line 233, in forward
core_attn_out = self._checkpointed_attention_forward(query, key, value, attention_mask)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/transformer/attention.py", line 81, in _checkpointed_attention_forward
hidden_states = tensor_parallel.checkpoint(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/tensor_parallel/random.py", line 249, in checkpoint
return CheckpointFunction.apply(function, distribute_saved_activations, *args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/torch/autograd/function.py", line 539, in apply
return super().apply(*args, **kwargs) # type: ignore[misc]
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/tensor_parallel/random.py", line 191, in forward
outputs = run_function(*args)
^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/megatron/core/transformer/attention.py", line 78, in custom_forward
output_ = self.dot_product_attention(query, key, value, attention_mask)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/transformer_engine/pytorch/attention.py", line 2085, in forward
qkv_layout = _get_qkv_layout(query_layer, key_layer, value_layer,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/mambaforge/envs/nemo/lib/python3.11/site-packages/transformer_engine/pytorch/attention.py", line 1195, in _get_qkv_layout
raise Exception("The provided qkv memory layout is not supported!")
Exception: The provided qkv memory layout is not supported!
```
**Expected behavior**
Train a model properly.
**Environment overview (please complete the following information)**
- Environment location: Bare-metal
- Method of NeMo install:
```
$ pip install torch==2.1.0 torchvision --index-url https://download.pytorch.org/whl/cu118
$ git clone https://github.com/NVIDIA/apex
$ pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./apex
$ pip install -r requirements.txt # except torch
$ pip install -r requirements_lightning
$ pip install -r requirements_common # except youtokentome, run manually by conda install youtokentome
$ pip install -r requirements_nlp.txt
```
**Environment details**
If NVIDIA docker image is used you don't need to specify these.
Otherwise, please provide:
- OS version: ubuntu 20.04
- PyTorch version: 2.1.0
- Python version: 3.11
**Additional context**
GPU: A100 1node 8gpus
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7899/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/7899/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7897 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7897/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7897/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7897/events | https://github.com/NVIDIA/NeMo/pull/7897 | 1,996,144,718 | PR_kwDOC_bI7s5fl3J4 | 7,897 | FSDP + Tensor Parallelism | {
"login": "erhoo82",
"id": 15861259,
"node_id": "MDQ6VXNlcjE1ODYxMjU5",
"avatar_url": "https://avatars.githubusercontent.com/u/15861259?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/erhoo82",
"html_url": "https://github.com/erhoo82",
"followers_url": "https://api.github.com/users/erhoo82/followers",
"following_url": "https://api.github.com/users/erhoo82/following{/other_user}",
"gists_url": "https://api.github.com/users/erhoo82/gists{/gist_id}",
"starred_url": "https://api.github.com/users/erhoo82/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/erhoo82/subscriptions",
"organizations_url": "https://api.github.com/users/erhoo82/orgs",
"repos_url": "https://api.github.com/users/erhoo82/repos",
"events_url": "https://api.github.com/users/erhoo82/events{/privacy}",
"received_events_url": "https://api.github.com/users/erhoo82/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"@erhoo82 Thanks for hard work. Do you have timeline in mind when the PR will get merged? Thanks!",
"@stu1130 Have been caught up by high priority work. Trying to close soon.",
"@athitten, @ericharper \r\n- I have removed the torch FSDP patch file as the bug was fixed at ... | 2023-11-16T06:19:06 | 2024-01-02T01:44:58 | 2023-12-16T01:22:30 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7897",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7897",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7897.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7897.patch",
"merged_at": "2023-12-16T01:22:30"
} | # What does this PR do ?
This supports torch FSDP training with Tensor-parallelism.
Set `model.fsdp=True` to enable FSDP. Gradient reduction data type can be set as `model.fsdp_grad_reduce_dtype=bf16`.
FSDP mode is not compatible with the Distributed Optimizer and O2 optimizer wrapper, thus should set `model.megatron_amp_O2=False` and not use `model.optim.name=distributed_fused_adam`.
To store checkpoint in FSDP sharded format, set model.fsdp_sharded_checkpoint=True.
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7897/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/7897/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7896 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7896/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7896/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7896/events | https://github.com/NVIDIA/NeMo/pull/7896 | 1,995,780,997 | PR_kwDOC_bI7s5fkpy1 | 7,896 | [ASR] Online Speaker Diarization with ASR: Part-1 | {
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.github.com/users/tango4j/followers",
"following_url": "https://api.github.com/users/tango4j/following{/other_user}",
"gists_url": "https://api.github.com/users/tango4j/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tango4j/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tango4j/subscriptions",
"organizations_url": "https://api.github.com/users/tango4j/orgs",
"repos_url": "https://api.github.com/users/tango4j/repos",
"events_url": "https://api.github.com/users/tango4j/events{/privacy}",
"received_events_url": "https://api.github.com/users/tango4j/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "htt... | closed | false | {
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.github.com/users/tango4j/followers",
"following_url": "https://api.github.com/users/tango4j/following{/other_user}",
"gists_url": "https://api.github.com/users/tango4j/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tango4j/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tango4j/subscriptions",
"organizations_url": "https://api.github.com/users/tango4j/orgs",
"repos_url": "https://api.github.com/users/tango4j/repos",
"events_url": "https://api.github.com/users/tango4j/events{/privacy}",
"received_events_url": "https://api.github.com/users/tango4j/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.git... | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days."
] | 2023-11-15T23:58:57 | 2023-12-06T21:26:45 | 2023-12-06T21:26:44 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7896",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7896",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7896.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7896.patch",
"merged_at": null
} | # What does this PR do ?
This PR addresses utility functions (not core functions) for online diarization with ASR (multi-speaker ASR).
Streaming ASR and diarization utility functions and classes are added with unit tests.
This is Part1 of the clustering online diarization integration.
In Part2, example file, yaml file and tutorial notebook will be uploaded.
**Collection**: ASR
# Changelog
`nemo/collections/asr/models/__init__.py`
Adding `OnlineClusteringDiarizer` class
`nemo/collections/asr/parts/utils/decoder_timestamps_utils.py`
adding `FrameBatchASRLogitsSample` and `FrameBatchASRLogits`
for ASR with timestamp generation
`nemo/collections/asr/parts/utils/diarization_utils.py`
`OnlineDiarWithASR` and a few utility functions
`nemo/collections/asr/parts/utils/streaming_utils.py`
Adding `FrameBatchVAD` class which handles VAD streaming.
# Usage
Usage will be included in part2 with example file and yaml file.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [x] Did you add or update any necessary documentation?
- [] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in NeMo ASR.
# Additional Information
N.A. | {
"login": "tango4j",
"id": 13211483,
"node_id": "MDQ6VXNlcjEzMjExNDgz",
"avatar_url": "https://avatars.githubusercontent.com/u/13211483?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tango4j",
"html_url": "https://github.com/tango4j",
"followers_url": "https://api.github.com/users/tango4j/followers",
"following_url": "https://api.github.com/users/tango4j/following{/other_user}",
"gists_url": "https://api.github.com/users/tango4j/gists{/gist_id}",
"starred_url": "https://api.github.com/users/tango4j/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/tango4j/subscriptions",
"organizations_url": "https://api.github.com/users/tango4j/orgs",
"repos_url": "https://api.github.com/users/tango4j/repos",
"events_url": "https://api.github.com/users/tango4j/events{/privacy}",
"received_events_url": "https://api.github.com/users/tango4j/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7896/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/7896/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7895 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7895/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7895/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7895/events | https://github.com/NVIDIA/NeMo/pull/7895 | 1,995,717,611 | PR_kwDOC_bI7s5fkb-g | 7,895 | Add support for non-FP32 distributed optimizer state | {
"login": "timmoon10",
"id": 4406448,
"node_id": "MDQ6VXNlcjQ0MDY0NDg=",
"avatar_url": "https://avatars.githubusercontent.com/u/4406448?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/timmoon10",
"html_url": "https://github.com/timmoon10",
"followers_url": "https://api.github.com/users/timmoon10/followers",
"following_url": "https://api.github.com/users/timmoon10/following{/other_user}",
"gists_url": "https://api.github.com/users/timmoon10/gists{/gist_id}",
"starred_url": "https://api.github.com/users/timmoon10/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/timmoon10/subscriptions",
"organizations_url": "https://api.github.com/users/timmoon10/orgs",
"repos_url": "https://api.github.com/users/timmoon10/repos",
"events_url": "https://api.github.com/users/timmoon10/events{/privacy}",
"received_events_url": "https://api.github.com/users/timmoon10/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-15T22:54:33 | 2023-12-09T01:45:16 | 2023-12-09T01:45:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7895",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7895",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7895.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7895.patch",
"merged_at": null
} | # What does this PR do ?
The distributed Adam optimizer can now be configured with BF16 or FP16 optimizer state.
**Collection**: NLP
# Changelog
- Add support for non-FP32 distributed optimizer state
# Usage
Run a language model, e.g. with the config at https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/language_modeling/conf/megatron_gpt_config.yaml.
Enable the distributed optimizer with `model.optim.name=distributed_fused_adam` and set the optimizer state dtype with `+model.optim.dtype=bfloat16`.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7895/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/7895/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7894 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7894/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7894/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7894/events | https://github.com/NVIDIA/NeMo/pull/7894 | 1,995,610,718 | PR_kwDOC_bI7s5fkEij | 7,894 | Change FP8 Defaults | {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-15T21:41:16 | 2023-11-16T21:16:10 | 2023-11-16T21:16:09 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7894",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7894",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7894.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7894.patch",
"merged_at": "2023-11-16T21:16:09"
} | # What does this PR do ?
Change the certain FP8 parameter defaults to more recommended values so that new users can find the optimal FP8 recipe faster.
**Collection**: NLP
# Changelog
- Change all defaults of `fp8_amax_history_len` from 1 to 1024
- Change all defaults of `fp8_amax_compute_algo` from 'most_recent' to 'max'
- Change config defaults of `fp8_hybrid` to True to mark that hybrid is the preferred fp8 mode (note `fp8` is still False so by default FP8 is still off.)
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7894/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/7894/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7892 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7892/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7892/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7892/events | https://github.com/NVIDIA/NeMo/pull/7892 | 1,995,307,084 | PR_kwDOC_bI7s5fjB6b | 7,892 | add Dutch P&C FC model info | {
"login": "zhehuaichen",
"id": 139396994,
"node_id": "U_kgDOCE8Hgg",
"avatar_url": "https://avatars.githubusercontent.com/u/139396994?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhehuaichen",
"html_url": "https://github.com/zhehuaichen",
"followers_url": "https://api.github.com/users/zhehuaichen/followers",
"following_url": "https://api.github.com/users/zhehuaichen/following{/other_user}",
"gists_url": "https://api.github.com/users/zhehuaichen/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zhehuaichen/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zhehuaichen/subscriptions",
"organizations_url": "https://api.github.com/users/zhehuaichen/orgs",
"repos_url": "https://api.github.com/users/zhehuaichen/repos",
"events_url": "https://api.github.com/users/zhehuaichen/events{/privacy}",
"received_events_url": "https://api.github.com/users/zhehuaichen/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-15T18:27:31 | 2023-11-23T01:35:18 | 2023-11-23T01:35:17 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7892",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7892",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7892.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7892.patch",
"merged_at": "2023-11-23T01:35:17"
} | # What does this PR do ?
Update scores table for Fast Conformer Dutch models
**Collection**: [ASR]
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
| {
"login": "zhehuaichen",
"id": 139396994,
"node_id": "U_kgDOCE8Hgg",
"avatar_url": "https://avatars.githubusercontent.com/u/139396994?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhehuaichen",
"html_url": "https://github.com/zhehuaichen",
"followers_url": "https://api.github.com/users/zhehuaichen/followers",
"following_url": "https://api.github.com/users/zhehuaichen/following{/other_user}",
"gists_url": "https://api.github.com/users/zhehuaichen/gists{/gist_id}",
"starred_url": "https://api.github.com/users/zhehuaichen/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/zhehuaichen/subscriptions",
"organizations_url": "https://api.github.com/users/zhehuaichen/orgs",
"repos_url": "https://api.github.com/users/zhehuaichen/repos",
"events_url": "https://api.github.com/users/zhehuaichen/events{/privacy}",
"received_events_url": "https://api.github.com/users/zhehuaichen/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7892/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/7892/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7891 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7891/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7891/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7891/events | https://github.com/NVIDIA/NeMo/pull/7891 | 1,995,270,771 | PR_kwDOC_bI7s5fi6Fp | 7,891 | FSDP + Tensor Parallelism with Patch | {
"login": "erhoo82",
"id": 15861259,
"node_id": "MDQ6VXNlcjE1ODYxMjU5",
"avatar_url": "https://avatars.githubusercontent.com/u/15861259?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/erhoo82",
"html_url": "https://github.com/erhoo82",
"followers_url": "https://api.github.com/users/erhoo82/followers",
"following_url": "https://api.github.com/users/erhoo82/following{/other_user}",
"gists_url": "https://api.github.com/users/erhoo82/gists{/gist_id}",
"starred_url": "https://api.github.com/users/erhoo82/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/erhoo82/subscriptions",
"organizations_url": "https://api.github.com/users/erhoo82/orgs",
"repos_url": "https://api.github.com/users/erhoo82/repos",
"events_url": "https://api.github.com/users/erhoo82/events{/privacy}",
"received_events_url": "https://api.github.com/users/erhoo82/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-15T18:01:13 | 2023-11-16T04:45:28 | 2023-11-16T04:45:28 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7891",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7891",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7891.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7891.patch",
"merged_at": null
} | # What does this PR do ?
This supports torch FSDP training with Tensor-parallelism.
Set `model.fsdp=True` to enable FSDP. Gradient reduction data type can be set as `model.fsdp_grad_reduce_dtype=bf16`.
FSDP mode is not compatible with the Distributed Optimizer and O2 optimizer wrapper, thus should set `model.megatron_amp_O2=False` and not use `model.optim.name=distributed_fused_adam`.
To store checkpoint in FSDP sharded format, set `model.fsdp_sharded_checkpoint=True`.
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "erhoo82",
"id": 15861259,
"node_id": "MDQ6VXNlcjE1ODYxMjU5",
"avatar_url": "https://avatars.githubusercontent.com/u/15861259?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/erhoo82",
"html_url": "https://github.com/erhoo82",
"followers_url": "https://api.github.com/users/erhoo82/followers",
"following_url": "https://api.github.com/users/erhoo82/following{/other_user}",
"gists_url": "https://api.github.com/users/erhoo82/gists{/gist_id}",
"starred_url": "https://api.github.com/users/erhoo82/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/erhoo82/subscriptions",
"organizations_url": "https://api.github.com/users/erhoo82/orgs",
"repos_url": "https://api.github.com/users/erhoo82/repos",
"events_url": "https://api.github.com/users/erhoo82/events{/privacy}",
"received_events_url": "https://api.github.com/users/erhoo82/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7891/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/7891/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7890 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7890/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7890/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7890/events | https://github.com/NVIDIA/NeMo/pull/7890 | 1,994,697,917 | PR_kwDOC_bI7s5fg8w3 | 7,890 | Add support for swiglu fusion | {
"login": "lhb8125",
"id": 16619280,
"node_id": "MDQ6VXNlcjE2NjE5Mjgw",
"avatar_url": "https://avatars.githubusercontent.com/u/16619280?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lhb8125",
"html_url": "https://github.com/lhb8125",
"followers_url": "https://api.github.com/users/lhb8125/followers",
"following_url": "https://api.github.com/users/lhb8125/following{/other_user}",
"gists_url": "https://api.github.com/users/lhb8125/gists{/gist_id}",
"starred_url": "https://api.github.com/users/lhb8125/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/lhb8125/subscriptions",
"organizations_url": "https://api.github.com/users/lhb8125/orgs",
"repos_url": "https://api.github.com/users/lhb8125/repos",
"events_url": "https://api.github.com/users/lhb8125/events{/privacy}",
"received_events_url": "https://api.github.com/users/lhb8125/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
... | closed | false | null | [] | null | [] | 2023-11-15T12:37:59 | 2023-11-17T08:50:41 | 2023-11-17T08:50:41 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7890",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7890",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7890.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7890.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "lhb8125",
"id": 16619280,
"node_id": "MDQ6VXNlcjE2NjE5Mjgw",
"avatar_url": "https://avatars.githubusercontent.com/u/16619280?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/lhb8125",
"html_url": "https://github.com/lhb8125",
"followers_url": "https://api.github.com/users/lhb8125/followers",
"following_url": "https://api.github.com/users/lhb8125/following{/other_user}",
"gists_url": "https://api.github.com/users/lhb8125/gists{/gist_id}",
"starred_url": "https://api.github.com/users/lhb8125/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/lhb8125/subscriptions",
"organizations_url": "https://api.github.com/users/lhb8125/orgs",
"repos_url": "https://api.github.com/users/lhb8125/repos",
"events_url": "https://api.github.com/users/lhb8125/events{/privacy}",
"received_events_url": "https://api.github.com/users/lhb8125/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7890/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/7890/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7889 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7889/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7889/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7889/events | https://github.com/NVIDIA/NeMo/pull/7889 | 1,993,825,227 | PR_kwDOC_bI7s5fd_ow | 7,889 | Fix CPU initialization of GPT models | {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-15T00:50:35 | 2023-11-21T16:31:17 | 2023-11-21T16:31:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7889",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7889",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7889.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7889.patch",
"merged_at": "2023-11-21T16:31:16"
} | # What does this PR do ?
Add `use_cpu_initialization` check in MegatronGPTModel.
Allows conversion scripts such as `convert_nemo_gpt_to_mcore.py` to run for bigger models in cpu mode.
**Collection**: NLP
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7889/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/7889/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7888 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7888/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7888/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7888/events | https://github.com/NVIDIA/NeMo/pull/7888 | 1,993,154,986 | PR_kwDOC_bI7s5fbsYx | 7,888 | Distributed checkpoint averaging supports bf16 type | {
"login": "yidong72",
"id": 43824965,
"node_id": "MDQ6VXNlcjQzODI0OTY1",
"avatar_url": "https://avatars.githubusercontent.com/u/43824965?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yidong72",
"html_url": "https://github.com/yidong72",
"followers_url": "https://api.github.com/users/yidong72/followers",
"following_url": "https://api.github.com/users/yidong72/following{/other_user}",
"gists_url": "https://api.github.com/users/yidong72/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yidong72/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yidong72/subscriptions",
"organizations_url": "https://api.github.com/users/yidong72/orgs",
"repos_url": "https://api.github.com/users/yidong72/repos",
"events_url": "https://api.github.com/users/yidong72/events{/privacy}",
"received_events_url": "https://api.github.com/users/yidong72/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"jenkins"
] | 2023-11-14T16:56:05 | 2023-12-01T21:16:17 | 2023-12-01T21:16:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7888",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7888",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7888.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7888.patch",
"merged_at": "2023-12-01T21:16:16"
} | Add support of bf16 type. | {
"login": "yidong72",
"id": 43824965,
"node_id": "MDQ6VXNlcjQzODI0OTY1",
"avatar_url": "https://avatars.githubusercontent.com/u/43824965?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yidong72",
"html_url": "https://github.com/yidong72",
"followers_url": "https://api.github.com/users/yidong72/followers",
"following_url": "https://api.github.com/users/yidong72/following{/other_user}",
"gists_url": "https://api.github.com/users/yidong72/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yidong72/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yidong72/subscriptions",
"organizations_url": "https://api.github.com/users/yidong72/orgs",
"repos_url": "https://api.github.com/users/yidong72/repos",
"events_url": "https://api.github.com/users/yidong72/events{/privacy}",
"received_events_url": "https://api.github.com/users/yidong72/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7888/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/7888/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7887 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7887/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7887/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7887/events | https://github.com/NVIDIA/NeMo/pull/7887 | 1,992,714,031 | PR_kwDOC_bI7s5faMQt | 7,887 | Fix k2 installation: update for latest PyTorch, move script to `installers` dir | {
"login": "artbataev",
"id": 9380560,
"node_id": "MDQ6VXNlcjkzODA1NjA=",
"avatar_url": "https://avatars.githubusercontent.com/u/9380560?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/artbataev",
"html_url": "https://github.com/artbataev",
"followers_url": "https://api.github.com/users/artbataev/followers",
"following_url": "https://api.github.com/users/artbataev/following{/other_user}",
"gists_url": "https://api.github.com/users/artbataev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/artbataev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/artbataev/subscriptions",
"organizations_url": "https://api.github.com/users/artbataev/orgs",
"repos_url": "https://api.github.com/users/artbataev/repos",
"events_url": "https://api.github.com/users/artbataev/events{/privacy}",
"received_events_url": "https://api.github.com/users/artbataev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/CI",
"name": "CI",
"color": "1D76DB",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-14T13:08:57 | 2023-11-14T21:30:03 | 2023-11-14T21:02:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7887",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7887",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7887.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7887.patch",
"merged_at": "2023-11-14T21:02:16"
} | # What does this PR do ?
- fix k2 installation: use latest release compatible with the latest PyTorch version
- move k2 installer to `scripts/installers`directory
**Collection**: [CI, ASR]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [x] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7887/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/7887/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7886 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7886/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7886/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7886/events | https://github.com/NVIDIA/NeMo/pull/7886 | 1,991,832,810 | PR_kwDOC_bI7s5fXOVa | 7,886 | [Codec] Finite scalar quantizer | {
"login": "anteju",
"id": 108555623,
"node_id": "U_kgDOBnhtZw",
"avatar_url": "https://avatars.githubusercontent.com/u/108555623?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anteju",
"html_url": "https://github.com/anteju",
"followers_url": "https://api.github.com/users/anteju/followers",
"following_url": "https://api.github.com/users/anteju/following{/other_user}",
"gists_url": "https://api.github.com/users/anteju/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anteju/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anteju/subscriptions",
"organizations_url": "https://api.github.com/users/anteju/orgs",
"repos_url": "https://api.github.com/users/anteju/repos",
"events_url": "https://api.github.com/users/anteju/events{/privacy}",
"received_events_url": "https://api.github.com/users/anteju/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-14T02:00:56 | 2023-11-17T22:35:44 | 2023-11-17T22:35:44 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7886",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7886",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7886.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7886.patch",
"merged_at": "2023-11-17T22:35:44"
} | # What does this PR do ?
This PR adds finite scalar quantizer as used in
Mentzer et al., Finite Scalar Quantization: VQ-VAE Made Simple (https://arxiv.org/abs/2309.15505v1)
**Collection**: TTS
# Changelog
- Added `FiniteScalarQuantizer`
- Added unit tests
# Usage
This can be used as a drop-in replacement for `ResidualVectorQuantizer`, for example using the following configuration
```yaml
model:
...
commit_loss_scale: 0.0 # need to be set to zero, since FSQ does not include commit loss
...
vector_quantizer:
_target_: nemo.collections.tts.modules.audio_codec_modules.FiniteScalarQuantizer
num_levels: [8, 5, 5, 5]
```
For using larger vector, `GroupFiniteScalarQuantizer` can be used to repeat quantization pattern several times, for example using the following configuration
```yaml
model:
...
commit_loss_scale: 0.0 # need to be set to zero, since FSQ does not include commit loss
...
vector_quantizer:
_target_: nemo.collections.tts.modules.audio_codec_modules.GroupFiniteScalarQuantizer
num_groups: 3
num_levels_per_group: [3, 4]
```
This corresponds to FSQ with `num_levels: [3, 4, 3, 4, 3, 4]`.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "anteju",
"id": 108555623,
"node_id": "U_kgDOBnhtZw",
"avatar_url": "https://avatars.githubusercontent.com/u/108555623?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anteju",
"html_url": "https://github.com/anteju",
"followers_url": "https://api.github.com/users/anteju/followers",
"following_url": "https://api.github.com/users/anteju/following{/other_user}",
"gists_url": "https://api.github.com/users/anteju/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anteju/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anteju/subscriptions",
"organizations_url": "https://api.github.com/users/anteju/orgs",
"repos_url": "https://api.github.com/users/anteju/repos",
"events_url": "https://api.github.com/users/anteju/events{/privacy}",
"received_events_url": "https://api.github.com/users/anteju/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7886/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/7886/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7885 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7885/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7885/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7885/events | https://github.com/NVIDIA/NeMo/pull/7885 | 1,991,715,650 | PR_kwDOC_bI7s5fW0QC | 7,885 | Distributed optimizer support for experimental FP8 tensors | {
"login": "timmoon10",
"id": 4406448,
"node_id": "MDQ6VXNlcjQ0MDY0NDg=",
"avatar_url": "https://avatars.githubusercontent.com/u/4406448?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/timmoon10",
"html_url": "https://github.com/timmoon10",
"followers_url": "https://api.github.com/users/timmoon10/followers",
"following_url": "https://api.github.com/users/timmoon10/following{/other_user}",
"gists_url": "https://api.github.com/users/timmoon10/gists{/gist_id}",
"starred_url": "https://api.github.com/users/timmoon10/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/timmoon10/subscriptions",
"organizations_url": "https://api.github.com/users/timmoon10/orgs",
"repos_url": "https://api.github.com/users/timmoon10/repos",
"events_url": "https://api.github.com/users/timmoon10/events{/privacy}",
"received_events_url": "https://api.github.com/users/timmoon10/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-14T00:31:38 | 2023-12-07T01:45:57 | 2023-12-07T01:45:57 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7885",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7885",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7885.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7885.patch",
"merged_at": null
} | # What does this PR do ?
This PR integrates with Transformer Engine's experimental `Float8Tensor`s. This allows the model to only store FP8 weight matrices. The distributed optimizer stores an FP32 master copy of the weights and performs param all-gathers in FP8.
**Collection**: NLP
# Changelog
- Adds distributed optimizer support for FP8 parameters
- Adds the option to initialize GPT with FP8 parameters
# Usage
Run GPT, e.g. with the config at https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/language_modeling/conf/megatron_gpt_config.yaml.
Enable FP8 support with `model.fp8=True`, FP8 parameters with `model.fp8_params=True`, and the distributed optimizer with `model.optim.name=distributed_fused_adam`.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [x] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [x] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Depends on https://github.com/NVIDIA/apex/pull/1749
* Uses Transformer Engine `Float8Tensor` added in https://github.com/NVIDIA/TransformerEngine/pull/452
* Closes https://github.com/NVIDIA/NeMo/pull/7469 and https://github.com/NVIDIA/NeMo/pull/7565 | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7885/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/7885/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7884 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7884/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7884/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7884/events | https://github.com/NVIDIA/NeMo/pull/7884 | 1,991,691,733 | PR_kwDOC_bI7s5fWvAR | 7,884 | [TTS] Add period discriminator and feature matching loss to codec recipe | {
"login": "rlangman",
"id": 9942053,
"node_id": "MDQ6VXNlcjk5NDIwNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/9942053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rlangman",
"html_url": "https://github.com/rlangman",
"followers_url": "https://api.github.com/users/rlangman/followers",
"following_url": "https://api.github.com/users/rlangman/following{/other_user}",
"gists_url": "https://api.github.com/users/rlangman/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rlangman/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rlangman/subscriptions",
"organizations_url": "https://api.github.com/users/rlangman/orgs",
"repos_url": "https://api.github.com/users/rlangman/repos",
"events_url": "https://api.github.com/users/rlangman/events{/privacy}",
"received_events_url": "https://api.github.com/users/rlangman/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was c... | 2023-11-14T00:12:30 | 2024-01-19T17:02:55 | 2024-01-19T17:02:54 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7884",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7884",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7884.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7884.patch",
"merged_at": "2024-01-19T17:02:54"
} | # What does this PR do ?
Add the feature matching loss and period discriminator used in most traditional vocoders. These will be used in my next PR which adds a configuration for creating a codec from mel spectrogram.
**Collection**: [TTS]
# Changelog
- Add absolute feature matching loss.
- Add period discriminator
- Add wrapper for supplying multiple discriminators.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation | {
"login": "rlangman",
"id": 9942053,
"node_id": "MDQ6VXNlcjk5NDIwNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/9942053?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rlangman",
"html_url": "https://github.com/rlangman",
"followers_url": "https://api.github.com/users/rlangman/followers",
"following_url": "https://api.github.com/users/rlangman/following{/other_user}",
"gists_url": "https://api.github.com/users/rlangman/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rlangman/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rlangman/subscriptions",
"organizations_url": "https://api.github.com/users/rlangman/orgs",
"repos_url": "https://api.github.com/users/rlangman/repos",
"events_url": "https://api.github.com/users/rlangman/events{/privacy}",
"received_events_url": "https://api.github.com/users/rlangman/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7884/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/7884/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7883 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7883/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7883/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7883/events | https://github.com/NVIDIA/NeMo/pull/7883 | 1,991,684,589 | PR_kwDOC_bI7s5fWtfj | 7,883 | different truncation types for long context truncation, addition of longbench tasks, preprocessing bug fixs | {
"login": "paul-gibbons",
"id": 87940629,
"node_id": "MDQ6VXNlcjg3OTQwNjI5",
"avatar_url": "https://avatars.githubusercontent.com/u/87940629?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/paul-gibbons",
"html_url": "https://github.com/paul-gibbons",
"followers_url": "https://api.github.com/users/paul-gibbons/followers",
"following_url": "https://api.github.com/users/paul-gibbons/following{/other_user}",
"gists_url": "https://api.github.com/users/paul-gibbons/gists{/gist_id}",
"starred_url": "https://api.github.com/users/paul-gibbons/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/paul-gibbons/subscriptions",
"organizations_url": "https://api.github.com/users/paul-gibbons/orgs",
"repos_url": "https://api.github.com/users/paul-gibbons/repos",
"events_url": "https://api.github.com/users/paul-gibbons/events{/privacy}",
"received_events_url": "https://api.github.com/users/paul-gibbons/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-14T00:03:45 | 2023-11-20T20:22:29 | 2023-11-20T20:22:29 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7883",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7883",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7883.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7883.patch",
"merged_at": "2023-11-20T20:22:29"
} | # What does this PR do ?
Adding able to truncate from left/middle/right portions of context.
Adding LongBench tasks to dataset preparation script.
Changed placement of remove_newline_tab flag in prepare_truncated_data.py so there these characters are removed prior to checking if the context's sequence length is greater than the max sequence length.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "shan18",
"id": 12576178,
"node_id": "MDQ6VXNlcjEyNTc2MTc4",
"avatar_url": "https://avatars.githubusercontent.com/u/12576178?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/shan18",
"html_url": "https://github.com/shan18",
"followers_url": "https://api.github.com/users/shan18/followers",
"following_url": "https://api.github.com/users/shan18/following{/other_user}",
"gists_url": "https://api.github.com/users/shan18/gists{/gist_id}",
"starred_url": "https://api.github.com/users/shan18/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/shan18/subscriptions",
"organizations_url": "https://api.github.com/users/shan18/orgs",
"repos_url": "https://api.github.com/users/shan18/repos",
"events_url": "https://api.github.com/users/shan18/events{/privacy}",
"received_events_url": "https://api.github.com/users/shan18/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7883/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/7883/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7882 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7882/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7882/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7882/events | https://github.com/NVIDIA/NeMo/pull/7882 | 1,991,567,722 | PR_kwDOC_bI7s5fWUL5 | 7,882 | Add back import guard | {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
... | closed | false | null | [] | null | [
"jenkins",
"jenkins"
] | 2023-11-13T22:08:19 | 2023-11-15T03:17:00 | 2023-11-15T03:16:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7882",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7882",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7882.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7882.patch",
"merged_at": "2023-11-15T03:16:59"
} | # What does this PR do ?
Add back the import guard that was deleted in a previous PR
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to #7750
| {
"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/7882/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/7882/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7881 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7881/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7881/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7881/events | https://github.com/NVIDIA/NeMo/pull/7881 | 1,991,351,452 | PR_kwDOC_bI7s5fVklP | 7,881 | add comment on script and fix target check | {
"login": "gshennvm",
"id": 119401249,
"node_id": "U_kgDOBx3rIQ",
"avatar_url": "https://avatars.githubusercontent.com/u/119401249?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gshennvm",
"html_url": "https://github.com/gshennvm",
"followers_url": "https://api.github.com/users/gshennvm/followers",
"following_url": "https://api.github.com/users/gshennvm/following{/other_user}",
"gists_url": "https://api.github.com/users/gshennvm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/gshennvm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/gshennvm/subscriptions",
"organizations_url": "https://api.github.com/users/gshennvm/orgs",
"repos_url": "https://api.github.com/users/gshennvm/repos",
"events_url": "https://api.github.com/users/gshennvm/events{/privacy}",
"received_events_url": "https://api.github.com/users/gshennvm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2023-11-13T19:41:22 | 2023-11-14T06:28:17 | 2023-11-14T06:28:16 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7881",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7881",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7881.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7881.patch",
"merged_at": "2023-11-14T06:28:16"
} | # What does this PR do ?
fix issue with target checking in nemo -> mcore conversion
**PR Type**:
- [x] Bugfix
| {
"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/7881/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/7881/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7880 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7880/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7880/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7880/events | https://github.com/NVIDIA/NeMo/pull/7880 | 1,991,218,429 | PR_kwDOC_bI7s5fVHF_ | 7,880 | NeMo + Lhotse integration | {
"login": "pzelasko",
"id": 15930688,
"node_id": "MDQ6VXNlcjE1OTMwNjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/15930688?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pzelasko",
"html_url": "https://github.com/pzelasko",
"followers_url": "https://api.github.com/users/pzelasko/followers",
"following_url": "https://api.github.com/users/pzelasko/following{/other_user}",
"gists_url": "https://api.github.com/users/pzelasko/gists{/gist_id}",
"starred_url": "https://api.github.com/users/pzelasko/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/pzelasko/subscriptions",
"organizations_url": "https://api.github.com/users/pzelasko/orgs",
"repos_url": "https://api.github.com/users/pzelasko/repos",
"events_url": "https://api.github.com/users/pzelasko/events{/privacy}",
"received_events_url": "https://api.github.com/users/pzelasko/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"@titu1994 all suggestions implemented",
"Need to pass jenkins though",
"jenkins",
"jenkins",
"Thanks guys! Such a joy to click merge after all ... | 2023-11-13T18:27:50 | 2024-01-13T00:24:14 | 2024-01-13T00:24:08 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7880",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7880",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7880.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7880.patch",
"merged_at": "2024-01-13T00:24:08"
} | # What does this PR do ?
This PR adds an option to leverage Lhotse for dataloading in NeMo.
**Collection**: Currently ASR only, with clear path to extend into other collections.
# Changelog
- Support creating a Lhotse dataloader from NeMo configuration
- Support for most Lhotse-specific dataloading features, especially: dynamic bucketing, multiplexing datasets.
- Support initializing Lhotse dataloader either from Lhotse manifests, Lhotse Shar (tarred) manifests, NeMo manifests, and NeMo tarred manifests
- ASR dataloading for hybrid CTC transducer as the first example
# Usage
* See the unit tests
* NeMo's `model.{train,validation,test}_ds` sections can be extended with additional fields `use_lhotse: True/False` and `lhotse: {...}` to enable this.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [x] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [x] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "pzelasko",
"id": 15930688,
"node_id": "MDQ6VXNlcjE1OTMwNjg4",
"avatar_url": "https://avatars.githubusercontent.com/u/15930688?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pzelasko",
"html_url": "https://github.com/pzelasko",
"followers_url": "https://api.github.com/users/pzelasko/followers",
"following_url": "https://api.github.com/users/pzelasko/following{/other_user}",
"gists_url": "https://api.github.com/users/pzelasko/gists{/gist_id}",
"starred_url": "https://api.github.com/users/pzelasko/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/pzelasko/subscriptions",
"organizations_url": "https://api.github.com/users/pzelasko/orgs",
"repos_url": "https://api.github.com/users/pzelasko/repos",
"events_url": "https://api.github.com/users/pzelasko/events{/privacy}",
"received_events_url": "https://api.github.com/users/pzelasko/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7880/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/7880/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7879 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7879/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7879/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7879/events | https://github.com/NVIDIA/NeMo/pull/7879 | 1,990,986,772 | PR_kwDOC_bI7s5fUUS8 | 7,879 | add values to en tts dict | {
"login": "mgrafu",
"id": 47233618,
"node_id": "MDQ6VXNlcjQ3MjMzNjE4",
"avatar_url": "https://avatars.githubusercontent.com/u/47233618?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mgrafu",
"html_url": "https://github.com/mgrafu",
"followers_url": "https://api.github.com/users/mgrafu/followers",
"following_url": "https://api.github.com/users/mgrafu/following{/other_user}",
"gists_url": "https://api.github.com/users/mgrafu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mgrafu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mgrafu/subscriptions",
"organizations_url": "https://api.github.com/users/mgrafu/orgs",
"repos_url": "https://api.github.com/users/mgrafu/repos",
"events_url": "https://api.github.com/users/mgrafu/events{/privacy}",
"received_events_url": "https://api.github.com/users/mgrafu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"@ekmb any updates on this?",
"> @ekmb any updates on this?\r\n\r\n@mgrafu I'm good with merge once the CI is green",
"This PR is stale because it h... | 2023-11-13T16:14:58 | 2024-01-30T19:01:47 | 2024-01-30T19:01:45 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7879",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7879",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7879.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7879.patch",
"merged_at": "2024-01-30T19:01:45"
} | # What does this PR do ?
Add terms to tts ipa pronunciation dictionary for en-US to meet internal needs OOTB
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "mgrafu",
"id": 47233618,
"node_id": "MDQ6VXNlcjQ3MjMzNjE4",
"avatar_url": "https://avatars.githubusercontent.com/u/47233618?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mgrafu",
"html_url": "https://github.com/mgrafu",
"followers_url": "https://api.github.com/users/mgrafu/followers",
"following_url": "https://api.github.com/users/mgrafu/following{/other_user}",
"gists_url": "https://api.github.com/users/mgrafu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mgrafu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mgrafu/subscriptions",
"organizations_url": "https://api.github.com/users/mgrafu/orgs",
"repos_url": "https://api.github.com/users/mgrafu/repos",
"events_url": "https://api.github.com/users/mgrafu/events{/privacy}",
"received_events_url": "https://api.github.com/users/mgrafu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7879/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/7879/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7878 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7878/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7878/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7878/events | https://github.com/NVIDIA/NeMo/issues/7878 | 1,988,837,751 | I_kwDOC_bI7s52i0F3 | 7,878 | Cannot resume fro checkpoint | {
"login": "githukelvin",
"id": 78278735,
"node_id": "MDQ6VXNlcjc4Mjc4NzM1",
"avatar_url": "https://avatars.githubusercontent.com/u/78278735?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/githukelvin",
"html_url": "https://github.com/githukelvin",
"followers_url": "https://api.github.com/users/githukelvin/followers",
"following_url": "https://api.github.com/users/githukelvin/following{/other_user}",
"gists_url": "https://api.github.com/users/githukelvin/gists{/gist_id}",
"starred_url": "https://api.github.com/users/githukelvin/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/githukelvin/subscriptions",
"organizations_url": "https://api.github.com/users/githukelvin/orgs",
"repos_url": "https://api.github.com/users/githukelvin/repos",
"events_url": "https://api.github.com/users/githukelvin/events{/privacy}",
"received_events_url": "https://api.github.com/users/githukelvin/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"In these cases, use a predefined version name for exp_manager, as you mentioned `exp_manager.resume_ignore_no_checkpoint=false` which means it would error out if there are no checkpoints in save directory. \r\nuse `exp_manager.version=v1` or something like that. ",
"it is not working though\r\n",
"what do... | 2023-11-11T08:30:16 | 2024-04-16T11:30:16 | 2023-12-24T01:46:28 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | i am training conformer ctc so i was in epoch 3 epoch three ended but it didnt save the checkpoint for that epoch rerunning again it is not getting any checkpoint neither 2 or one how can i resolve this it is starting again from zero it i saying no checkpoint found
<code>
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
`Trainer(val_check_interval=1.0)` was configured so validation will run at the end of the training epoch..
[NeMo W 2023-11-11 08:05:28 exp_manager:754] No version folders would be created under the log folder as 'resume_if_exists' is enabled.
Error executing job with overrides: ['exp_manager.name=SwahiliModelTraining', 'exp_manager.resume_if_exists=true', 'exp_manager.resume_ignore_no_checkpoint=false', 'exp_manager.exp_dir=drive/MyDrive/hackathon/', 'model.tokenizer.dir=drive/MyDrive/hackathon/tokenizer_spe_bpe_v1024_max_4', 'model.train_ds.is_tarred=true', 'model.train_ds.tarred_audio_filepaths=drive/MyDrive/hackathon/train_tarred_1bk/audio__OP_0..1023_CL_.tar', 'model.train_ds.manifest_filepath=drive/MyDrive/hackathon/train_tarred_1bk/tarred_audio_manifest.json', 'model.validation_ds.manifest_filepath=/content/drive/MyDrive/hackathon/eval_decoded_processed.json', 'model.test_ds.manifest_filepath=drive/MyDrive/hackathon/test_decoded_processed_v2.json']
Traceback (most recent call last):
File "/content/drive/MyDrive/hackathon/scripts/speech_to_text_ctc_bpe.py", line 81, in main
exp_manager(trainer, cfg.get("exp_manager", None))
File "/usr/local/lib/python3.10/dist-packages/nemo/utils/exp_manager.py", line 356, in exp_manager
check_resume(
File "/usr/local/lib/python3.10/dist-packages/nemo/utils/exp_manager.py", line 613, in check_resume
raise NotFoundError(
nemo.utils.exp_manager.NotFoundError: There were no checkpoints found in checkpoint_dir or no checkpoint folder at checkpoint_dir :drive/MyDrive/hackathon/SwahiliModelTraining/checkpoints. Cannot resume.
Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
</code>
as you can see above but the model when i run interference it is working fine
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7878/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/7878/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7877 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7877/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7877/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7877/events | https://github.com/NVIDIA/NeMo/issues/7877 | 1,988,666,235 | I_kwDOC_bI7s52iKN7 | 7,877 | Model card for SteerLM has incorrect command-line arguments | {
"login": "ndronen",
"id": 21194,
"node_id": "MDQ6VXNlcjIxMTk0",
"avatar_url": "https://avatars.githubusercontent.com/u/21194?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ndronen",
"html_url": "https://github.com/ndronen",
"followers_url": "https://api.github.com/users/ndronen/followers",
"following_url": "https://api.github.com/users/ndronen/following{/other_user}",
"gists_url": "https://api.github.com/users/ndronen/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ndronen/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ndronen/subscriptions",
"organizations_url": "https://api.github.com/users/ndronen/orgs",
"repos_url": "https://api.github.com/users/ndronen/repos",
"events_url": "https://api.github.com/users/ndronen/events{/privacy}",
"received_events_url": "https://api.github.com/users/ndronen/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-11T02:35:11 | 2023-12-20T01:39:58 | 2023-12-20T01:39:58 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
The instructions for launching the Steer-LM eval server in the HuggingFace model card are incorrect.
**Steps/Code to reproduce bug**
1. Follow [instructions to install prerequisites](https://huggingface.co/nvidia/SteerLM-llama2-13B).
2. Download [`LLAMA2-13B-SteerLM.nemo`](https://huggingface.co/nvidia/SteerLM-llama2-13B/tree/main).
3. Run the documented command: `python megatron_gpt_eval.py gpt_model_file=LLAMA2-13B-SteerLM.nemo trainer.precision=16 server=True tensor_model_parallel_size=4 trainer.devices=1 pipeline_model_parallel_split_rank=0`
**Expected behavior**
The expected behavior is that the eval server starts or, if the system resources are insufficient, an error occurs.
Ideally, the model card will say how to run the eval server depending on the available GPU memory and number of GPUs. I'd like to be able to run this on a 4xV100 machine.
Instead, I see the following:
```
python megatron_gpt_eval.py gpt_model_file=LLAMA2-13B-SteerLM.nemo trainer.precision=16 server=True tensor_model_parallel_size=4 trainer.devices=1 pipeline_model_parallel_split_rank=0
[NeMo W 2023-11-11 02:15:43 experimental:27] Module <class 'nemo.collections.nlp.data.language_modeling.megatron.megatron_batch_samplers.MegatronPretrainingRandomBatchSampler'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2023-11-11 02:15:43 experimental:27] Module <class 'nemo.collections.nlp.models.text_normalization_as_tagging.thutmose_tagger.ThutmoseTaggerModel'> is experimental, not ready for production and is not fully supported. Use at your own risk.
[NeMo W 2023-11-11 02:15:48 nemo_logging:349] /home/ubuntu/venv/lib/python3.10/site-packages/hydra/_internal/hydra.py:119: UserWarning: Future Hydra versions will no longer change working directory at job runtime by default.
See https://hydra.cc/docs/next/upgrades/1.1_to_1.2/changes_to_job_working_dir/ for more information.
ret = run_job(
Using 16bit None Automatic Mixed Precision (AMP)
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
IPU available: False, using: 0 IPUs
HPU available: False, using: 0 HPUs
Error executing job with overrides: ['gpt_model_file=LLAMA2-13B-SteerLM.nemo', 'trainer.precision=16', 'server=True', 'tensor_model_parallel_size=4', 'trainer.devices=1', 'pipeline_model_parallel_split_rank=0']
Traceback (most recent call last):
File "/home/ubuntu/NeMo/examples/nlp/language_modeling/megatron_gpt_eval.py", line 161, in main
cfg.trainer.devices * cfg.trainer.num_nodes
AssertionError: devices * num_nodes should equal tensor_model_parallel_size * pipeline_model_parallel_size
```
**Environment overview**
- Environment location: AWS EC2
- Method of NeMo install: `pip install -e .` in `HEAD detached at v1.17.0`
- If method of install is [Docker], provide `docker pull` & `docker run` commands used
**Environment details**
```
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 22.04.3 LTS
Release: 22.04
Codename: jammy
```
```
>>> torch.__version__
'2.1.0+cu121'
```
```
python -V
Python 3.10.12
```
**Additional context**
Example: 4xV100
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7877/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/7877/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7875 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7875/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7875/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7875/events | https://github.com/NVIDIA/NeMo/pull/7875 | 1,986,068,212 | PR_kwDOC_bI7s5fDvj4 | 7,875 | Mengruw/fixed tgt lang v3 | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
},
{
"id": 4556669155,
"node_id": "LA_kwDOC_bI7s8AAAABD5lA4w",
"url": "https://a... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-09T17:11:56 | 2023-12-01T01:49:46 | 2023-12-01T01:49:46 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7875",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7875",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7875.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7875.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7875/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/7875/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7874 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7874/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7874/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7874/events | https://github.com/NVIDIA/NeMo/issues/7874 | 1,985,905,525 | I_kwDOC_bI7s52XoN1 | 7,874 | swig_decoders dependency | {
"login": "vsokolovskii",
"id": 48914918,
"node_id": "MDQ6VXNlcjQ4OTE0OTE4",
"avatar_url": "https://avatars.githubusercontent.com/u/48914918?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsokolovskii",
"html_url": "https://github.com/vsokolovskii",
"followers_url": "https://api.github.com/users/vsokolovskii/followers",
"following_url": "https://api.github.com/users/vsokolovskii/following{/other_user}",
"gists_url": "https://api.github.com/users/vsokolovskii/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsokolovskii/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsokolovskii/subscriptions",
"organizations_url": "https://api.github.com/users/vsokolovskii/orgs",
"repos_url": "https://api.github.com/users/vsokolovskii/repos",
"events_url": "https://api.github.com/users/vsokolovskii/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsokolovskii/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"
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
... | closed | false | null | [] | null | [
"It should have been installed as per [line](https://github.com/NVIDIA/NeMo/blob/77d1386174930239fb619a7d4660a88f2061a5b3/scripts/installers/setup_os2s_decoders.py#L138C18-L138C30).\r\n\r\n\r\nWas it unsuccessful during installation? Can you check your logs and post the error here? ",
"This issue is stale because... | 2023-11-09T15:42:56 | 2024-10-13T09:42:28 | 2023-12-19T01:45:56 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
Cannot run the n-gram beam search script.
**Steps/Code to reproduce bug**
Run the `./install_beamsearch_decoders.sh`
``` import swig_decoders
ModuleNotFoundError: No module named 'swig_decoders'
```
Ubuntu 22.04.3
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7874/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/7874/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7873 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7873/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7873/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7873/events | https://github.com/NVIDIA/NeMo/issues/7873 | 1,985,401,218 | I_kwDOC_bI7s52VtGC | 7,873 | NeMo MCore Llama2 support fp8 | {
"login": "chiendb97",
"id": 25007718,
"node_id": "MDQ6VXNlcjI1MDA3NzE4",
"avatar_url": "https://avatars.githubusercontent.com/u/25007718?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/chiendb97",
"html_url": "https://github.com/chiendb97",
"followers_url": "https://api.github.com/users/chiendb97/followers",
"following_url": "https://api.github.com/users/chiendb97/following{/other_user}",
"gists_url": "https://api.github.com/users/chiendb97/gists{/gist_id}",
"starred_url": "https://api.github.com/users/chiendb97/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/chiendb97/subscriptions",
"organizations_url": "https://api.github.com/users/chiendb97/orgs",
"repos_url": "https://api.github.com/users/chiendb97/repos",
"events_url": "https://api.github.com/users/chiendb97/events{/privacy}",
"received_events_url": "https://api.github.com/users/chiendb97/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | 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://a... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-09T11:14:12 | 2023-12-18T01:46:03 | 2023-12-18T01:46:03 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null |
Hi, does Nemo MCore llama2 currently support training with fp8? I got this error when training the llama model with fp8
Thanks ! | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7873/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/7873/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7872 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7872/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7872/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7872/events | https://github.com/NVIDIA/NeMo/pull/7872 | 1,984,257,911 | PR_kwDOC_bI7s5e9lXW | 7,872 | Add MCore T5 | {
"login": "cuichenx",
"id": 43478052,
"node_id": "MDQ6VXNlcjQzNDc4MDUy",
"avatar_url": "https://avatars.githubusercontent.com/u/43478052?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cuichenx",
"html_url": "https://github.com/cuichenx",
"followers_url": "https://api.github.com/users/cuichenx/followers",
"following_url": "https://api.github.com/users/cuichenx/following{/other_user}",
"gists_url": "https://api.github.com/users/cuichenx/gists{/gist_id}",
"starred_url": "https://api.github.com/users/cuichenx/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/cuichenx/subscriptions",
"organizations_url": "https://api.github.com/users/cuichenx/orgs",
"repos_url": "https://api.github.com/users/cuichenx/repos",
"events_url": "https://api.github.com/users/cuichenx/events{/privacy}",
"received_events_url": "https://api.github.com/users/cuichenx/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-08T19:26:38 | 2023-12-06T01:45:57 | 2023-12-06T01:45:56 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7872",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7872",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7872.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7872.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7872/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/7872/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7871 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7871/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7871/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7871/events | https://github.com/NVIDIA/NeMo/issues/7871 | 1,983,074,345 | I_kwDOC_bI7s52M1Ap | 7,871 | GPU usage by NEMO ASR Model | {
"login": "Oheed911",
"id": 83069712,
"node_id": "MDQ6VXNlcjgzMDY5NzEy",
"avatar_url": "https://avatars.githubusercontent.com/u/83069712?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Oheed911",
"html_url": "https://github.com/Oheed911",
"followers_url": "https://api.github.com/users/Oheed911/followers",
"following_url": "https://api.github.com/users/Oheed911/following{/other_user}",
"gists_url": "https://api.github.com/users/Oheed911/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Oheed911/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Oheed911/subscriptions",
"organizations_url": "https://api.github.com/users/Oheed911/orgs",
"repos_url": "https://api.github.com/users/Oheed911/repos",
"events_url": "https://api.github.com/users/Oheed911/events{/privacy}",
"received_events_url": "https://api.github.com/users/Oheed911/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"You can try using this before important anything from NeMo - https://github.com/pytorch/pytorch/blob/main/torch/cuda/memory.py#L121"
] | 2023-11-08T08:56:41 | 2023-11-28T00:46:37 | 2023-11-28T00:46:37 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I want to limit the usage of Nemo[Asr] mode, lets say i want to limit it to 5 GB of VRAM. how can i do it? is there any way to limit this? | {
"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/7871/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/7871/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7870 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7870/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7870/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7870/events | https://github.com/NVIDIA/NeMo/issues/7870 | 1,982,892,591 | I_kwDOC_bI7s52MIov | 7,870 | Issue faced in Nemo Model Inference using Farward collection | {
"login": "Pri4tam",
"id": 90524632,
"node_id": "MDQ6VXNlcjkwNTI0NjMy",
"avatar_url": "https://avatars.githubusercontent.com/u/90524632?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Pri4tam",
"html_url": "https://github.com/Pri4tam",
"followers_url": "https://api.github.com/users/Pri4tam/followers",
"following_url": "https://api.github.com/users/Pri4tam/following{/other_user}",
"gists_url": "https://api.github.com/users/Pri4tam/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Pri4tam/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Pri4tam/subscriptions",
"organizations_url": "https://api.github.com/users/Pri4tam/orgs",
"repos_url": "https://api.github.com/users/Pri4tam/repos",
"events_url": "https://api.github.com/users/Pri4tam/events{/privacy}",
"received_events_url": "https://api.github.com/users/Pri4tam/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:... | null | [
"Your reproducible steps are unclear as you are using some variables which are out of scope from what described. \r\n\r\nHowever for forward, signal should be BxT and signal_length should be B size **tensors**. \r\n\r\n Waveform.shape[idx] would return int but not torch tensor. \r\n\r\nCan you change it to `torch.t... | 2023-11-08T07:02:45 | 2023-11-28T00:48:51 | 2023-11-28T00:48:51 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
The farward nemo collection of Nemo ASR model requires following parameters.
- input_signal – Tensor that represents a batch of raw audio signals, of shape [B, T]. T here represents timesteps, with 1 second of audio represented as self.sample_rate number of floating point values.
- input_signal_length – Vector of length B, that contains the individual lengths of the audio sequences.
**Steps/Code to reproduce bug**
Load the model
hi_nemo_asr=nemo_asr.models.EncDecCTCModelBPE.restore_from(en_model_path).cpu()
waveform, sample_rate = torchaudio.load(speech_file[0])
input_signal_length=input_signal.shape[1]
log_prob,_,_=hi_nemo_asr.forward(input_signal=waveform,input_signal_length=waveform.shape[1])
This gives the following error
IndexError: tuple index out of range
**Environment details**
If NVIDIA docker image is used you don't need to specify these.
Otherwise, please provide:
-Os =Linux
- PyTorch version =2.1.0
- Python version=3.8.10
| {
"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/7870/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/7870/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7869 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7869/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7869/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7869/events | https://github.com/NVIDIA/NeMo/issues/7869 | 1,982,810,647 | I_kwDOC_bI7s52L0oX | 7,869 | Continue training from 'ckpt' file | {
"login": "TonSotin",
"id": 132354615,
"node_id": "U_kgDOB-OSNw",
"avatar_url": "https://avatars.githubusercontent.com/u/132354615?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/TonSotin",
"html_url": "https://github.com/TonSotin",
"followers_url": "https://api.github.com/users/TonSotin/followers",
"following_url": "https://api.github.com/users/TonSotin/following{/other_user}",
"gists_url": "https://api.github.com/users/TonSotin/gists{/gist_id}",
"starred_url": "https://api.github.com/users/TonSotin/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/TonSotin/subscriptions",
"organizations_url": "https://api.github.com/users/TonSotin/orgs",
"repos_url": "https://api.github.com/users/TonSotin/repos",
"events_url": "https://api.github.com/users/TonSotin/events{/privacy}",
"received_events_url": "https://api.github.com/users/TonSotin/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"if you are using `https://github.com/NVIDIA/NeMo/blob/main/examples/asr/speech_to_text_finetune.py` to finetune then we currently don;t support loading from ckpt file. \r\nHowever while training you can always save nemo file by toggling https://github.com/NVIDIA/NeMo/blob/d49b73cb495dfd17554b01cf1b0f6f8f1bbbbc7c/n... | 2023-11-08T06:02:37 | 2023-11-28T00:49:34 | 2023-11-28T00:49:34 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Is it possible somehow to continue training from checkpoint that made by pytorch_lightning? I have been training ASR model and load pretrained model. I want to continue training from exact position when server shut down. Do I need load from model.nemo or is there more gentle way to do this? | {
"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/7869/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/7869/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7868 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7868/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7868/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7868/events | https://github.com/NVIDIA/NeMo/pull/7868 | 1,980,367,954 | PR_kwDOC_bI7s5ewRB7 | 7,868 | added converter for lora checkpoints to tp1 | {
"login": "hutm",
"id": 699538,
"node_id": "MDQ6VXNlcjY5OTUzOA==",
"avatar_url": "https://avatars.githubusercontent.com/u/699538?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hutm",
"html_url": "https://github.com/hutm",
"followers_url": "https://api.github.com/users/hutm/followers",
"following_url": "https://api.github.com/users/hutm/following{/other_user}",
"gists_url": "https://api.github.com/users/hutm/gists{/gist_id}",
"starred_url": "https://api.github.com/users/hutm/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/hutm/subscriptions",
"organizations_url": "https://api.github.com/users/hutm/orgs",
"repos_url": "https://api.github.com/users/hutm/repos",
"events_url": "https://api.github.com/users/hutm/events{/privacy}",
"received_events_url": "https://api.github.com/users/hutm/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-07T01:30:49 | 2023-12-21T00:27:49 | 2023-11-29T01:45:55 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7868",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7868",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7868.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7868.patch",
"merged_at": null
} | # What does this PR do ?
added converter for lora checkpoints to tp1
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7868/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/7868/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7867 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7867/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7867/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7867/events | https://github.com/NVIDIA/NeMo/pull/7867 | 1,980,212,282 | PR_kwDOC_bI7s5evvT6 | 7,867 | Tar codec | {
"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
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-06T22:46:51 | 2023-11-18T02:47:11 | 2023-11-18T02:47:10 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7867",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7867",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7867.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7867.patch",
"merged_at": "2023-11-18T02:47:10"
} | Add webdataset support for VocoderDataset
**Collection**: TTS
# Changelog
- Add TarredVocderDataset class for dataset fetching
- sampler is not supported with IterableDataset
- added a conditional code to calculate correct number of max steps when webdataset is used
Known issue:
- Haven;t tested for passing multiple tarred datasets. Will be pushing a PR later to support it.
# Usage
* add `is_sharded=True` in cfg for train_ds to load webdataset
# Before your PR is Ready for review
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation | {
"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/7867/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/7867/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7866 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7866/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7866/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7866/events | https://github.com/NVIDIA/NeMo/pull/7866 | 1,980,149,682 | PR_kwDOC_bI7s5evhPg | 7,866 | Mengruw/fixed tgt lang v3 | {
"login": "mengruwNv",
"id": 149420466,
"node_id": "U_kgDOCOf5sg",
"avatar_url": "https://avatars.githubusercontent.com/u/149420466?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mengruwNv",
"html_url": "https://github.com/mengruwNv",
"followers_url": "https://api.github.com/users/mengruwNv/followers",
"following_url": "https://api.github.com/users/mengruwNv/following{/other_user}",
"gists_url": "https://api.github.com/users/mengruwNv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mengruwNv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mengruwNv/subscriptions",
"organizations_url": "https://api.github.com/users/mengruwNv/orgs",
"repos_url": "https://api.github.com/users/mengruwNv/repos",
"events_url": "https://api.github.com/users/mengruwNv/events{/privacy}",
"received_events_url": "https://api.github.com/users/mengruwNv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
},
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-06T22:05:12 | 2023-11-09T17:11:23 | 2023-11-09T17:11:22 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7866",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7866",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7866.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7866.patch",
"merged_at": "2023-11-09T17:11:22"
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7866/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/7866/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7865 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7865/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7865/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7865/events | https://github.com/NVIDIA/NeMo/pull/7865 | 1,980,142,959 | PR_kwDOC_bI7s5evfwH | 7,865 | Tar codec | {
"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
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
},
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://a... | closed | false | null | [] | null | [] | 2023-11-06T22:00:57 | 2023-11-06T22:44:38 | 2023-11-06T22:44:38 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7865",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7865",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7865.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7865.patch",
"merged_at": null
} | # What does this PR do ?
Add webdataset support for VocoderDataset
**Collection**: TTS
# Changelog
- Add TarredVocderDataset class for dataset fetching
- sampler is not supported with IterableDataset
- added a conditional code to calculate correct number of max steps when webdataset is used
# Usage
* add `is_sharded=True` in cfg for train_ds to load webdataset
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation | {
"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/7865/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/7865/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7864 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7864/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7864/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7864/events | https://github.com/NVIDIA/NeMo/pull/7864 | 1,980,010,097 | PR_kwDOC_bI7s5evCof | 7,864 | Removing duplicate Megatron-LM installation | {
"login": "Davood-M",
"id": 3964285,
"node_id": "MDQ6VXNlcjM5NjQyODU=",
"avatar_url": "https://avatars.githubusercontent.com/u/3964285?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Davood-M",
"html_url": "https://github.com/Davood-M",
"followers_url": "https://api.github.com/users/Davood-M/followers",
"following_url": "https://api.github.com/users/Davood-M/following{/other_user}",
"gists_url": "https://api.github.com/users/Davood-M/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Davood-M/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Davood-M/subscriptions",
"organizations_url": "https://api.github.com/users/Davood-M/orgs",
"repos_url": "https://api.github.com/users/Davood-M/repos",
"events_url": "https://api.github.com/users/Davood-M/events{/privacy}",
"received_events_url": "https://api.github.com/users/Davood-M/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/CI",
"name": "CI",
"color": "1D76DB",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [] | 2023-11-06T20:31:39 | 2023-11-07T02:08:36 | 2023-11-07T02:08:35 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7864",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7864",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7864.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7864.patch",
"merged_at": "2023-11-07T02:08:35"
} | # What does this PR do ?
small fix to remove duplicate megatron-lm installation
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7864/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/7864/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7863 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7863/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7863/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7863/events | https://github.com/NVIDIA/NeMo/pull/7863 | 1,979,669,089 | PR_kwDOC_bI7s5et2lz | 7,863 | Added comma so I can test build process. | {
"login": "DougAtNvidia",
"id": 150055323,
"node_id": "U_kgDOCPGpmw",
"avatar_url": "https://avatars.githubusercontent.com/u/150055323?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/DougAtNvidia",
"html_url": "https://github.com/DougAtNvidia",
"followers_url": "https://api.github.com/users/DougAtNvidia/followers",
"following_url": "https://api.github.com/users/DougAtNvidia/following{/other_user}",
"gists_url": "https://api.github.com/users/DougAtNvidia/gists{/gist_id}",
"starred_url": "https://api.github.com/users/DougAtNvidia/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/DougAtNvidia/subscriptions",
"organizations_url": "https://api.github.com/users/DougAtNvidia/orgs",
"repos_url": "https://api.github.com/users/DougAtNvidia/repos",
"events_url": "https://api.github.com/users/DougAtNvidia/events{/privacy}",
"received_events_url": "https://api.github.com/users/DougAtNvidia/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2023-11-06T17:25:57 | 2023-11-10T00:27:32 | 2023-11-10T00:27:32 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7863",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7863",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7863.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7863.patch",
"merged_at": null
} | # What does this PR do ?
Makes a trivial change.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [x] Did you add or update any necessary documentation?
- [x] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [x] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7863/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/7863/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7862 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7862/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7862/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7862/events | https://github.com/NVIDIA/NeMo/pull/7862 | 1,979,347,045 | PR_kwDOC_bI7s5esv5x | 7,862 | Upgrade NeMo to latest mcore and TE | {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
},
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://a... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-06T14:47:27 | 2023-11-17T13:11:02 | 2023-11-17T13:11:00 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7862",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7862",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7862.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7862.patch",
"merged_at": "2023-11-17T13:11:00"
} | # What does this PR do ?
Upgrades NeMo to latest mcore and TE versions.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7862/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/7862/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7861 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7861/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7861/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7861/events | https://github.com/NVIDIA/NeMo/issues/7861 | 1,977,997,600 | I_kwDOC_bI7s515dkg | 7,861 | 16000 sample rate VITS config help inquiry | {
"login": "ganymedenet",
"id": 51125243,
"node_id": "MDQ6VXNlcjUxMTI1MjQz",
"avatar_url": "https://avatars.githubusercontent.com/u/51125243?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ganymedenet",
"html_url": "https://github.com/ganymedenet",
"followers_url": "https://api.github.com/users/ganymedenet/followers",
"following_url": "https://api.github.com/users/ganymedenet/following{/other_user}",
"gists_url": "https://api.github.com/users/ganymedenet/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ganymedenet/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ganymedenet/subscriptions",
"organizations_url": "https://api.github.com/users/ganymedenet/orgs",
"repos_url": "https://api.github.com/users/ganymedenet/repos",
"events_url": "https://api.github.com/users/ganymedenet/events{/privacy}",
"received_events_url": "https://api.github.com/users/ganymedenet/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"
},
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
... | closed | false | {
"login": "XuesongYang",
"id": 1646669,
"node_id": "MDQ6VXNlcjE2NDY2Njk=",
"avatar_url": "https://avatars.githubusercontent.com/u/1646669?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/XuesongYang",
"html_url": "https://github.com/XuesongYang",
"followers_url": "https://api.github.com/users/XuesongYang/followers",
"following_url": "https://api.github.com/users/XuesongYang/following{/other_user}",
"gists_url": "https://api.github.com/users/XuesongYang/gists{/gist_id}",
"starred_url": "https://api.github.com/users/XuesongYang/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/XuesongYang/subscriptions",
"organizations_url": "https://api.github.com/users/XuesongYang/orgs",
"repos_url": "https://api.github.com/users/XuesongYang/repos",
"events_url": "https://api.github.com/users/XuesongYang/events{/privacy}",
"received_events_url": "https://api.github.com/users/XuesongYang/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "XuesongYang",
"id": 1646669,
"node_id": "MDQ6VXNlcjE2NDY2Njk=",
"avatar_url": "https://avatars.githubusercontent.com/u/1646669?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/XuesongYang",
"html_url": "https://github.com/XuesongYang",
"followers_url": "https... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale.",
"This issue is stale because it has been open for 30 days with no activity. ... | 2023-11-05T22:14:42 | 2024-03-04T01:46:54 | 2024-03-04T01:46:54 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Good day,
Could you help me to re-configure my VITS model trainer for 16000 hz speech dataset?
sample_rate: 22050
n_mel_channels: 80
n_window_size: 1024
n_window_stride: 256
n_fft: 1024
lowfreq: 0
highfreq: null
window: hann
I tried different combination but always receive something like these:
```
RuntimeError: The size of tensor a (64) must match the size of tensor b (128) at non-singleton dimension 2
RuntimeError: The size of tensor a (40) must match the size of tensor b (51) at non-singleton dimension 2
```
Thank you!
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7861/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/7861/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7860 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7860/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7860/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7860/events | https://github.com/NVIDIA/NeMo/issues/7860 | 1,977,456,799 | I_kwDOC_bI7s513Zif | 7,860 | Question: can we load same size bert model to train same size gpt if yes then how? | {
"login": "mujhenahiata",
"id": 94825377,
"node_id": "U_kgDOBabroQ",
"avatar_url": "https://avatars.githubusercontent.com/u/94825377?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mujhenahiata",
"html_url": "https://github.com/mujhenahiata",
"followers_url": "https://api.github.com/users/mujhenahiata/followers",
"following_url": "https://api.github.com/users/mujhenahiata/following{/other_user}",
"gists_url": "https://api.github.com/users/mujhenahiata/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mujhenahiata/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mujhenahiata/subscriptions",
"organizations_url": "https://api.github.com/users/mujhenahiata/orgs",
"repos_url": "https://api.github.com/users/mujhenahiata/repos",
"events_url": "https://api.github.com/users/mujhenahiata/events{/privacy}",
"received_events_url": "https://api.github.com/users/mujhenahiata/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | closed | false | {
"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
} | [
{
"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:... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed becau... | 2023-11-04T18:21:14 | 2024-01-13T01:45:51 | 2024-01-13T01:45:51 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I have trained a bert model, I want to use its weights to train gpt model. Is it possible, if yes how
@meghmak13
@titu1994
@nithinraok
@okuchaiev | {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7860/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/7860/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7859 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7859/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7859/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7859/events | https://github.com/NVIDIA/NeMo/pull/7859 | 1,977,215,529 | PR_kwDOC_bI7s5elxbb | 7,859 | Adding support for TTS generated SFT for speechlm models | {
"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
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4847373924,
"node_id": "LA_kwDOC_bI7s8AAAABIO0OZA",
"url": "htt... | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-04T07:33:27 | 2023-12-11T01:46:09 | 2023-12-11T01:46:08 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7859",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7859",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7859.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7859.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7859/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/7859/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7858 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7858/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7858/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7858/events | https://github.com/NVIDIA/NeMo/pull/7858 | 1,976,619,939 | PR_kwDOC_bI7s5ejyfD | 7,858 | generalized chat sft prompt (#7655) | {
"login": "dimapihtar",
"id": 37850217,
"node_id": "MDQ6VXNlcjM3ODUwMjE3",
"avatar_url": "https://avatars.githubusercontent.com/u/37850217?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dimapihtar",
"html_url": "https://github.com/dimapihtar",
"followers_url": "https://api.github.com/users/dimapihtar/followers",
"following_url": "https://api.github.com/users/dimapihtar/following{/other_user}",
"gists_url": "https://api.github.com/users/dimapihtar/gists{/gist_id}",
"starred_url": "https://api.github.com/users/dimapihtar/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/dimapihtar/subscriptions",
"organizations_url": "https://api.github.com/users/dimapihtar/orgs",
"repos_url": "https://api.github.com/users/dimapihtar/repos",
"events_url": "https://api.github.com/users/dimapihtar/events{/privacy}",
"received_events_url": "https://api.github.com/users/dimapihtar/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
},
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "htt... | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-03T17:13:48 | 2023-11-27T01:45:50 | 2023-11-27T01:45:50 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7858",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7858",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7858.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7858.patch",
"merged_at": null
} | * fix dataset issues
* working version
* all passed
* refactor tests
* all pass
* working version
* use end name signal for labels
* all fixed
* update doc
* style fix
* remove unused imports
* make sure nccl not timing out
* style fix
* generate example template
* generic end of name token
* style fix
* add the chat prompt format into the config
* make sure sft working
* address reviewer comment
* fix non
* try openAI prompt
* remove unused imports
* remove human labels from the data
* use hf dataset to clean
* reviewer comments
---------
# What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7858/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/7858/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7857 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7857/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7857/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7857/events | https://github.com/NVIDIA/NeMo/issues/7857 | 1,976,520,232 | I_kwDOC_bI7s51z04o | 7,857 | `preprocessor` in `exportables` but not exported | {
"login": "msis",
"id": 577139,
"node_id": "MDQ6VXNlcjU3NzEzOQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/577139?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/msis",
"html_url": "https://github.com/msis",
"followers_url": "https://api.github.com/users/msis/followers",
"following_url": "https://api.github.com/users/msis/following{/other_user}",
"gists_url": "https://api.github.com/users/msis/gists{/gist_id}",
"starred_url": "https://api.github.com/users/msis/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/msis/subscriptions",
"organizations_url": "https://api.github.com/users/msis/orgs",
"repos_url": "https://api.github.com/users/msis/repos",
"events_url": "https://api.github.com/users/msis/events{/privacy}",
"received_events_url": "https://api.github.com/users/msis/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 | [
"This is by (forced) design. Preprocessor is exportable in one very specific case, when you use the torchaudio backend of the preprocessor. Otherwise it is not exportable. There's no easy way to declare a class is exportable in exactly one case where it's instance is derived from a mirrored API in torchaudio, so we... | 2023-11-03T16:10:46 | 2023-11-08T14:43:54 | 2023-11-08T14:43:54 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
The `preprocessor` is an instance of `Exportable` and yet it is not exported with `model.export()`
**Steps/Code to reproduce bug**
```py
>> exportables = []
>> for m in model.modules():
if isinstance(m, Exportable):
exportables.append(m)
>> exportables
[EncDecCTCModelBPE(
(preprocessor): AudioToMelSpectrogramPreprocessor(
(featurizer): FilterbankFeaturesTA(
(_mel_spec_extractor): MelSpectrogram(
(spectrogram): Spectrogram()
(mel_scale): MelScale()
)
)
)
(encoder): ConformerEncoder(
(pre_encode): ConvSubsampling(
(out): Linear(in_features=10240, out_features=512, bias=True)
(conv): Sequential(
(0): Conv2d(1, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
(1): ReLU(inplace=True)
(2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
(3): ReLU(inplace=True)
)
)
(pos_enc): RelPositionalEncoding(
(dropout): Dropout(p=0.1, inplace=False)
)
(layers): ModuleList(
(0-17): 18 x ConformerLayer(
(norm_feed_forward1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(feed_forward1): ConformerFeedForward(
(linear1): Linear(in_features=512, out_features=2048, bias=True)
(activation): Swish()
(dropout): Dropout(p=0.1, inplace=False)
(linear2): Linear(in_features=2048, out_features=512, bias=True)
)
(norm_conv): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(conv): ConformerConvolution(
(pointwise_conv1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))
(depthwise_conv): CausalConv1D(512, 512, kernel_size=(31,), stride=(1,), groups=512)
(batch_norm): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(activation): Swish()
(pointwise_conv2): Conv1d(512, 512, kernel_size=(1,), stride=(1,))
)
(norm_self_att): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(self_attn): RelPositionMultiHeadAttention(
(linear_q): Linear(in_features=512, out_features=512, bias=True)
(linear_k): Linear(in_features=512, out_features=512, bias=True)
(linear_v): Linear(in_features=512, out_features=512, bias=True)
(linear_out): Linear(in_features=512, out_features=512, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(linear_pos): Linear(in_features=512, out_features=512, bias=False)
)
(norm_feed_forward2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(feed_forward2): ConformerFeedForward(
(linear1): Linear(in_features=512, out_features=2048, bias=True)
(activation): Swish()
(dropout): Dropout(p=0.1, inplace=False)
(linear2): Linear(in_features=2048, out_features=512, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
(norm_out): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
)
)
)
(decoder): ConvASRDecoder(
(decoder_layers): Sequential(
(0): Conv1d(512, 129, kernel_size=(1,), stride=(1,))
)
)
(loss): CTCLoss()
(spec_augmentation): SpectrogramAugmentation(
(spec_augment): SpecAugment()
)
(_wer): WERBPE()
),
AudioToMelSpectrogramPreprocessor(
(featurizer): FilterbankFeaturesTA(
(_mel_spec_extractor): MelSpectrogram(
(spectrogram): Spectrogram()
(mel_scale): MelScale()
)
)
),
ConformerEncoder(
(pre_encode): ConvSubsampling(
(out): Linear(in_features=10240, out_features=512, bias=True)
(conv): Sequential(
(0): Conv2d(1, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
(1): ReLU(inplace=True)
(2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
(3): ReLU(inplace=True)
)
)
(pos_enc): RelPositionalEncoding(
(dropout): Dropout(p=0.1, inplace=False)
)
(layers): ModuleList(
(0-17): 18 x ConformerLayer(
(norm_feed_forward1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(feed_forward1): ConformerFeedForward(
(linear1): Linear(in_features=512, out_features=2048, bias=True)
(activation): Swish()
(dropout): Dropout(p=0.1, inplace=False)
(linear2): Linear(in_features=2048, out_features=512, bias=True)
)
(norm_conv): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(conv): ConformerConvolution(
(pointwise_conv1): Conv1d(512, 1024, kernel_size=(1,), stride=(1,))
(depthwise_conv): CausalConv1D(512, 512, kernel_size=(31,), stride=(1,), groups=512)
(batch_norm): BatchNorm1d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(activation): Swish()
(pointwise_conv2): Conv1d(512, 512, kernel_size=(1,), stride=(1,))
)
(norm_self_att): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(self_attn): RelPositionMultiHeadAttention(
(linear_q): Linear(in_features=512, out_features=512, bias=True)
(linear_k): Linear(in_features=512, out_features=512, bias=True)
(linear_v): Linear(in_features=512, out_features=512, bias=True)
(linear_out): Linear(in_features=512, out_features=512, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(linear_pos): Linear(in_features=512, out_features=512, bias=False)
)
(norm_feed_forward2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
(feed_forward2): ConformerFeedForward(
(linear1): Linear(in_features=512, out_features=2048, bias=True)
(activation): Swish()
(dropout): Dropout(p=0.1, inplace=False)
(linear2): Linear(in_features=2048, out_features=512, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
(norm_out): LayerNorm((512,), eps=1e-05, elementwise_affine=True)
)
)
),
ConvASRDecoder(
(decoder_layers): Sequential(
(0): Conv1d(512, 129, kernel_size=(1,), stride=(1,))
)
)]
```
**Expected behavior**
```py
m, descriptions = model.export(MODEL_OUT)
```
should create a model that contains the preprocessor and the rest of the model.
However, the exported model skips the preprocessor
**Environment overview (please complete the following information)**
- Method of NeMo install: `pip install nemo_toolkit[all]`
**Environment details**
Otherwise, please provide:
- OS version: macOS 14.1
- PyTorch version: 3.10.13
- Python version: 2.1.0
**Additional context**
Trying to convert Conformer model to CoreML
| {
"login": "msis",
"id": 577139,
"node_id": "MDQ6VXNlcjU3NzEzOQ==",
"avatar_url": "https://avatars.githubusercontent.com/u/577139?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/msis",
"html_url": "https://github.com/msis",
"followers_url": "https://api.github.com/users/msis/followers",
"following_url": "https://api.github.com/users/msis/following{/other_user}",
"gists_url": "https://api.github.com/users/msis/gists{/gist_id}",
"starred_url": "https://api.github.com/users/msis/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/msis/subscriptions",
"organizations_url": "https://api.github.com/users/msis/orgs",
"repos_url": "https://api.github.com/users/msis/repos",
"events_url": "https://api.github.com/users/msis/events{/privacy}",
"received_events_url": "https://api.github.com/users/msis/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7857/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/7857/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7856 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7856/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7856/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7856/events | https://github.com/NVIDIA/NeMo/pull/7856 | 1,976,321,223 | PR_kwDOC_bI7s5ei0MF | 7,856 | Bugfix: Args didn't have spe byte fallback | {
"login": "AstralScribe",
"id": 105854650,
"node_id": "U_kgDOBk82ug",
"avatar_url": "https://avatars.githubusercontent.com/u/105854650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AstralScribe",
"html_url": "https://github.com/AstralScribe",
"followers_url": "https://api.github.com/users/AstralScribe/followers",
"following_url": "https://api.github.com/users/AstralScribe/following{/other_user}",
"gists_url": "https://api.github.com/users/AstralScribe/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AstralScribe/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AstralScribe/subscriptions",
"organizations_url": "https://api.github.com/users/AstralScribe/orgs",
"repos_url": "https://api.github.com/users/AstralScribe/repos",
"events_url": "https://api.github.com/users/AstralScribe/events{/privacy}",
"received_events_url": "https://api.github.com/users/AstralScribe/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | closed | false | null | [] | null | [
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-03T14:27:03 | 2023-12-02T01:44:42 | 2023-12-02T01:44:42 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7856",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7856",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7856.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7856.patch",
"merged_at": null
} | # What does this PR do ?
process_asr_text_tokenizer didn't have parser for args: spe_byte_fallback.
**Collection**: [ASR]
# Changelog
- Bugfix: Added spe_byte_fallback to the args list.
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
## Who can review?
@titu1994, @redoctopus, @jbalam-nv, or @okuchaiev
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7856/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/7856/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7855 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7855/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7855/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7855/events | https://github.com/NVIDIA/NeMo/issues/7855 | 1,975,342,977 | I_kwDOC_bI7s51vVeB | 7,855 | Runtime error | {
"login": "sriks8",
"id": 8636333,
"node_id": "MDQ6VXNlcjg2MzYzMzM=",
"avatar_url": "https://avatars.githubusercontent.com/u/8636333?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sriks8",
"html_url": "https://github.com/sriks8",
"followers_url": "https://api.github.com/users/sriks8/followers",
"following_url": "https://api.github.com/users/sriks8/following{/other_user}",
"gists_url": "https://api.github.com/users/sriks8/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sriks8/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sriks8/subscriptions",
"organizations_url": "https://api.github.com/users/sriks8/orgs",
"repos_url": "https://api.github.com/users/sriks8/repos",
"events_url": "https://api.github.com/users/sriks8/events{/privacy}",
"received_events_url": "https://api.github.com/users/sriks8/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Its an error from gradio not nemo. Not sure what you are trying to do. ",
"Thanks for looking at this. \r\n\r\nI tried to deploy the model (https://huggingface.co/nvidia/stt_en_fastconformer_transducer_xlarge) as an gardio app in click deploy option on the right) but it is failing. It errors out during the build... | 2023-11-03T02:01:14 | 2023-11-09T01:44:47 | 2023-11-09T01:44:46 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Traceback (most recent call last):
File "/home/user/app/app.py", line 3, in <module>
gr.Interface.load("models/nvidia/stt_en_fastconformer_transducer_xlarge").launch()
AttributeError: 'Dependency' object has no attribute 'launch' | {
"login": "sriks8",
"id": 8636333,
"node_id": "MDQ6VXNlcjg2MzYzMzM=",
"avatar_url": "https://avatars.githubusercontent.com/u/8636333?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sriks8",
"html_url": "https://github.com/sriks8",
"followers_url": "https://api.github.com/users/sriks8/followers",
"following_url": "https://api.github.com/users/sriks8/following{/other_user}",
"gists_url": "https://api.github.com/users/sriks8/gists{/gist_id}",
"starred_url": "https://api.github.com/users/sriks8/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/sriks8/subscriptions",
"organizations_url": "https://api.github.com/users/sriks8/orgs",
"repos_url": "https://api.github.com/users/sriks8/repos",
"events_url": "https://api.github.com/users/sriks8/events{/privacy}",
"received_events_url": "https://api.github.com/users/sriks8/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7855/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/7855/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7854 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7854/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7854/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7854/events | https://github.com/NVIDIA/NeMo/pull/7854 | 1,975,171,716 | PR_kwDOC_bI7s5ee4Wi | 7,854 | Update transformers cache on Jenkins | {
"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
} | [
{
"id": 4840216855,
"node_id": "LA_kwDOC_bI7s8AAAABIH_ZFw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/CI",
"name": "CI",
"color": "1D76DB",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"jenkins",
"jenkins"
] | 2023-11-02T23:00:36 | 2023-11-03T06:20:13 | 2023-11-03T06:20:12 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7854",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7854",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7854.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7854.patch",
"merged_at": "2023-11-03T06:20:12"
} | # What does this PR do ?
Updating the cache by allowing transformers to be online
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"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/7854/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/7854/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7853 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7853/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7853/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7853/events | https://github.com/NVIDIA/NeMo/pull/7853 | 1,974,980,532 | PR_kwDOC_bI7s5eeOBI | 7,853 | enable reading phonetic transcripts from manifest | {
"login": "mgrafu",
"id": 47233618,
"node_id": "MDQ6VXNlcjQ3MjMzNjE4",
"avatar_url": "https://avatars.githubusercontent.com/u/47233618?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mgrafu",
"html_url": "https://github.com/mgrafu",
"followers_url": "https://api.github.com/users/mgrafu/followers",
"following_url": "https://api.github.com/users/mgrafu/following{/other_user}",
"gists_url": "https://api.github.com/users/mgrafu/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mgrafu/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mgrafu/subscriptions",
"organizations_url": "https://api.github.com/users/mgrafu/orgs",
"repos_url": "https://api.github.com/users/mgrafu/repos",
"events_url": "https://api.github.com/users/mgrafu/events{/privacy}",
"received_events_url": "https://api.github.com/users/mgrafu/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/TTS",
"name": "TTS",
"color": "49AF28",
"default": false,
"description": ""
},
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://a... | closed | false | null | [] | null | [
"2 cleaner approaches to consider:\r\n\r\n- Define a G2P class which sets its symbols to `get_ipa_character_set(locale)` and with a forward method that returns `[c for c in text]`\r\n- Make the g2p class an optional input to the IPATokenizer. Replace the checks with `if read_phones_from_file` to instead check `if g... | 2023-11-02T20:26:48 | 2023-11-25T01:44:36 | 2023-11-25T01:44:35 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7853",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7853",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7853.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7853.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7853/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/7853/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7852 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7852/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7852/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7852/events | https://github.com/NVIDIA/NeMo/pull/7852 | 1,974,800,397 | PR_kwDOC_bI7s5ednJJ | 7,852 | Fix flash decoding precision | {
"login": "hsiehjackson",
"id": 37269846,
"node_id": "MDQ6VXNlcjM3MjY5ODQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/37269846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hsiehjackson",
"html_url": "https://github.com/hsiehjackson",
"followers_url": "https://api.github.com/users/hsiehjackson/followers",
"following_url": "https://api.github.com/users/hsiehjackson/following{/other_user}",
"gists_url": "https://api.github.com/users/hsiehjackson/gists{/gist_id}",
"starred_url": "https://api.github.com/users/hsiehjackson/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/hsiehjackson/subscriptions",
"organizations_url": "https://api.github.com/users/hsiehjackson/orgs",
"repos_url": "https://api.github.com/users/hsiehjackson/repos",
"events_url": "https://api.github.com/users/hsiehjackson/events{/privacy}",
"received_events_url": "https://api.github.com/users/hsiehjackson/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811886210,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ogg",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/NLP",
"name": "NLP",
"color": "CBF09A",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins"
] | 2023-11-02T18:07:11 | 2023-11-06T21:40:00 | 2023-11-06T21:39:59 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7852",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7852",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7852.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7852.patch",
"merged_at": "2023-11-06T21:39:59"
} | # What does this PR do ?
Query and Key may have fp32 after `apply_rotary_pos_emb`, so we cast the precision to fp16 or bf16.
**Collection**: [NLP]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "hsiehjackson",
"id": 37269846,
"node_id": "MDQ6VXNlcjM3MjY5ODQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/37269846?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/hsiehjackson",
"html_url": "https://github.com/hsiehjackson",
"followers_url": "https://api.github.com/users/hsiehjackson/followers",
"following_url": "https://api.github.com/users/hsiehjackson/following{/other_user}",
"gists_url": "https://api.github.com/users/hsiehjackson/gists{/gist_id}",
"starred_url": "https://api.github.com/users/hsiehjackson/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/hsiehjackson/subscriptions",
"organizations_url": "https://api.github.com/users/hsiehjackson/orgs",
"repos_url": "https://api.github.com/users/hsiehjackson/repos",
"events_url": "https://api.github.com/users/hsiehjackson/events{/privacy}",
"received_events_url": "https://api.github.com/users/hsiehjackson/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7852/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/7852/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7851 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7851/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7851/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7851/events | https://github.com/NVIDIA/NeMo/pull/7851 | 1,973,689,627 | PR_kwDOC_bI7s5eZ0PP | 7,851 | Mengruw/fixed tgt lang ids inference/v2 | {
"login": "mengruwNv",
"id": 149420466,
"node_id": "U_kgDOCOf5sg",
"avatar_url": "https://avatars.githubusercontent.com/u/149420466?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mengruwNv",
"html_url": "https://github.com/mengruwNv",
"followers_url": "https://api.github.com/users/mengruwNv/followers",
"following_url": "https://api.github.com/users/mengruwNv/following{/other_user}",
"gists_url": "https://api.github.com/users/mengruwNv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/mengruwNv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/mengruwNv/subscriptions",
"organizations_url": "https://api.github.com/users/mengruwNv/orgs",
"repos_url": "https://api.github.com/users/mengruwNv/repos",
"events_url": "https://api.github.com/users/mengruwNv/events{/privacy}",
"received_events_url": "https://api.github.com/users/mengruwNv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1923495043,
"node_id": "MDU6TGFiZWwxOTIzNDk1MDQz",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/core",
"name": "core",
"color": "f25e3a",
"default": false,
"description": "Changes to NeMo Core"
},
{
"id": 4418253970,
"node_id": "LA_kwDOC_bI7s8AAAABB1k0kg",
... | closed | false | null | [] | null | [
"Hi, @Davood-M ",
"This PR is stale because it has been open for 14 days with no activity. Remove stale label or comment or update or this will be closed in 7 days.",
"This PR was closed because it has been inactive for 7 days since being marked as stale."
] | 2023-11-02T07:47:41 | 2023-11-25T01:44:38 | 2023-11-25T01:44:37 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7851",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7851",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7851.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7851.patch",
"merged_at": null
} | # What does this PR do ?
Add a one line overview of what this PR aims to accomplish.
**Collection**: [Note which collection this PR will affect]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [ ] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7851/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/7851/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7850 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7850/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7850/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7850/events | https://github.com/NVIDIA/NeMo/issues/7850 | 1,973,681,643 | I_kwDOC_bI7s51o_3r | 7,850 | Does the project support QLora | {
"login": "2033329616",
"id": 26346306,
"node_id": "MDQ6VXNlcjI2MzQ2MzA2",
"avatar_url": "https://avatars.githubusercontent.com/u/26346306?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/2033329616",
"html_url": "https://github.com/2033329616",
"followers_url": "https://api.github.com/users/2033329616/followers",
"following_url": "https://api.github.com/users/2033329616/following{/other_user}",
"gists_url": "https://api.github.com/users/2033329616/gists{/gist_id}",
"starred_url": "https://api.github.com/users/2033329616/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/2033329616/subscriptions",
"organizations_url": "https://api.github.com/users/2033329616/orgs",
"repos_url": "https://api.github.com/users/2033329616/repos",
"events_url": "https://api.github.com/users/2033329616/events{/privacy}",
"received_events_url": "https://api.github.com/users/2033329616/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4585431311,
"node_id": "LA_kwDOC_bI7s8AAAABEVAhDw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/stale",
"name": "stale",
"color": "ededed",
"default": false,
"description": null
}
] | 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://a... | null | [
"This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.",
"This issue was closed because it has been inactive for 7 days since being marked as stale.",
"@minitu Is there any plan for QloRA support ? thanks."
] | 2023-11-02T07:41:31 | 2024-04-03T10:23:37 | 2023-12-10T01:47:56 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Is your feature request related to a problem? Please describe.**
I have limit gpu resources， so I want to use qlora with model paralelism for LLM model.
**Describe the solution you'd like**
Using qlora and model paralelism or pipeline parallelism together。
**Describe alternatives you've considered**
None
**Additional context**
None
| {
"login": "github-actions[bot]",
"id": 41898282,
"node_id": "MDM6Qm90NDE4OTgyODI=",
"avatar_url": "https://avatars.githubusercontent.com/in/15368?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/github-actions%5Bbot%5D",
"html_url": "https://github.com/apps/github-actions",
"followers_url": "https://api.github.com/users/github-actions%5Bbot%5D/followers",
"following_url": "https://api.github.com/users/github-actions%5Bbot%5D/following{/other_user}",
"gists_url": "https://api.github.com/users/github-actions%5Bbot%5D/gists{/gist_id}",
"starred_url": "https://api.github.com/users/github-actions%5Bbot%5D/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/github-actions%5Bbot%5D/subscriptions",
"organizations_url": "https://api.github.com/users/github-actions%5Bbot%5D/orgs",
"repos_url": "https://api.github.com/users/github-actions%5Bbot%5D/repos",
"events_url": "https://api.github.com/users/github-actions%5Bbot%5D/events{/privacy}",
"received_events_url": "https://api.github.com/users/github-actions%5Bbot%5D/received_events",
"type": "Bot",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7850/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/7850/timeline | null | not_planned | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/7849 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7849/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7849/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7849/events | https://github.com/NVIDIA/NeMo/pull/7849 | 1,973,598,670 | PR_kwDOC_bI7s5eZg2K | 7,849 | [ASR] GSS-based mask estimator | {
"login": "anteju",
"id": 108555623,
"node_id": "U_kgDOBnhtZw",
"avatar_url": "https://avatars.githubusercontent.com/u/108555623?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anteju",
"html_url": "https://github.com/anteju",
"followers_url": "https://api.github.com/users/anteju/followers",
"following_url": "https://api.github.com/users/anteju/following{/other_user}",
"gists_url": "https://api.github.com/users/anteju/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anteju/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anteju/subscriptions",
"organizations_url": "https://api.github.com/users/anteju/orgs",
"repos_url": "https://api.github.com/users/anteju/repos",
"events_url": "https://api.github.com/users/anteju/events{/privacy}",
"received_events_url": "https://api.github.com/users/anteju/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | [
"jenkins",
"jenkins",
"jenkins",
"`ImportError while loading conftest '/home/taejinp/projects/gss_mask/NeMo/tests/conftest.py'.\r\ntests/conftest.py:28: in <module>\r\n from tests.fixtures.tts import *\r\ntests/fixtures/tts.py:21: in <module>\r\n from nemo.collections.asr.parts.utils.manifest_utils impo... | 2023-11-02T06:32:53 | 2023-11-16T00:56:29 | 2023-11-16T00:56:29 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7849",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7849",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7849.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7849.patch",
"merged_at": "2023-11-16T00:56:29"
} | # What does this PR do ?
Adding GSS mask estimator for use in multispeaker scenarios.
**Collection**: ASR
# Changelog
- Added an implementation of mask estimator using directional statistics clustering and activity guidance
- Added a unit test to test the module can be initialized and the output has the expected shape
# Usage
```python
import torch
from nemo.collections.asr.modules.audio_modules import MaskEstimatorGSS
batch_size, num_channels, num_subbands, num_frames = 1, 4, 257, 100
num_outputs = 3
# input (mixture) spectrogram
spec = torch.randn(batch_size, num_channels, num_subbands, num_frames, dtype=torch.cfloat)
# estimated source activity
source_activity = torch.randn(batch_size, num_outputs, num_frames) > 0.5
# mask estimator
me = MaskEstimatorGSS(num_iterations=5)
# estimated source masks
masks = me(input=spec, activity=source_activity)
assert masks.shape == (batch_size, num_outputs, num_subbands, num_frames)
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [x] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [ ] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [x] New Feature
- [ ] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "anteju",
"id": 108555623,
"node_id": "U_kgDOBnhtZw",
"avatar_url": "https://avatars.githubusercontent.com/u/108555623?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anteju",
"html_url": "https://github.com/anteju",
"followers_url": "https://api.github.com/users/anteju/followers",
"following_url": "https://api.github.com/users/anteju/following{/other_user}",
"gists_url": "https://api.github.com/users/anteju/gists{/gist_id}",
"starred_url": "https://api.github.com/users/anteju/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/anteju/subscriptions",
"organizations_url": "https://api.github.com/users/anteju/orgs",
"repos_url": "https://api.github.com/users/anteju/repos",
"events_url": "https://api.github.com/users/anteju/events{/privacy}",
"received_events_url": "https://api.github.com/users/anteju/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7849/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/7849/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/7848 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/7848/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/7848/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/7848/events | https://github.com/NVIDIA/NeMo/pull/7848 | 1,973,487,205 | PR_kwDOC_bI7s5eZJDo | 7,848 | Fix bug order of buckets not shuffled every epoch when using synced_r… | {
"login": "Alicu96",
"id": 112738869,
"node_id": "U_kgDOBrhCNQ",
"avatar_url": "https://avatars.githubusercontent.com/u/112738869?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Alicu96",
"html_url": "https://github.com/Alicu96",
"followers_url": "https://api.github.com/users/Alicu96/followers",
"following_url": "https://api.github.com/users/Alicu96/following{/other_user}",
"gists_url": "https://api.github.com/users/Alicu96/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Alicu96/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Alicu96/subscriptions",
"organizations_url": "https://api.github.com/users/Alicu96/orgs",
"repos_url": "https://api.github.com/users/Alicu96/repos",
"events_url": "https://api.github.com/users/Alicu96/events{/privacy}",
"received_events_url": "https://api.github.com/users/Alicu96/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 4811884691,
"node_id": "LA_kwDOC_bI7s8AAAABHs-Ikw",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/ASR",
"name": "ASR",
"color": "F3B33E",
"default": false,
"description": ""
}
] | closed | false | null | [] | null | null | 2023-11-02T04:32:57 | 2023-11-03T06:44:45 | 2023-11-03T06:44:35 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/7848",
"html_url": "https://github.com/NVIDIA/NeMo/pull/7848",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/7848.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/7848.patch",
"merged_at": null
} | …andomized
# What does this PR do ?
fix bug order of buckets not shuffled every epoch when using synced_randomized
**Collection**: [ASR]
# Changelog
- Add specific line by line info of high level changes in this PR.
# Usage
* You can potentially add a usage example below
```python
# Add a code snippet demonstrating how to use this
```
# Before your PR is "Ready for review"
**Pre checks**:
- [x] Make sure you read and followed [Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md)
- [ ] Did you write any new necessary tests?
- [ ] Did you add or update any necessary documentation?
- [x] Does the PR affect components that are optional to install? (Ex: Numba, Pynini, Apex etc)
- [ ] Reviewer: Does the PR have correct import guards for all optional libraries?
**PR Type**:
- [ ] New Feature
- [x] Bugfix
- [ ] Documentation
If you haven't finished some of the above items you can still open "Draft" PR.
## Who can review?
Anyone in the community is free to review the PR once the checks have passed.
[Contributor guidelines](https://github.com/NVIDIA/NeMo/blob/main/CONTRIBUTING.md) contains specific people who can review PRs to various areas.
# Additional Information
* Related to # (issue)
| {
"login": "Alicu96",
"id": 112738869,
"node_id": "U_kgDOBrhCNQ",
"avatar_url": "https://avatars.githubusercontent.com/u/112738869?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Alicu96",
"html_url": "https://github.com/Alicu96",
"followers_url": "https://api.github.com/users/Alicu96/followers",
"following_url": "https://api.github.com/users/Alicu96/following{/other_user}",
"gists_url": "https://api.github.com/users/Alicu96/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Alicu96/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Alicu96/subscriptions",
"organizations_url": "https://api.github.com/users/Alicu96/orgs",
"repos_url": "https://api.github.com/users/Alicu96/repos",
"events_url": "https://api.github.com/users/Alicu96/events{/privacy}",
"received_events_url": "https://api.github.com/users/Alicu96/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/7848/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/7848/timeline | null | null | true |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.