Datasets:
url
stringlengths 52
53
| repository_url
stringclasses 1
value | labels_url
stringlengths 66
67
| comments_url
stringlengths 61
62
| events_url
stringlengths 59
60
| html_url
stringlengths 40
43
| id
int64 719M
2.8B
| node_id
stringlengths 18
32
| number
int64 1.28k
11.9k
| title
stringlengths 1
350
| user
dict | labels
listlengths 0
10
| state
stringclasses 2
values | locked
bool 2
classes | assignee
dict | assignees
listlengths 0
3
| milestone
null | comments
sequencelengths 0
30
⌀ | created_at
timestamp[s] | updated_at
timestamp[s] | closed_at
timestamp[s] | author_association
stringclasses 4
values | sub_issues_summary
dict | active_lock_reason
stringclasses 1
value | draft
bool 2
classes | pull_request
dict | body
stringlengths 0
73.4k
⌀ | closed_by
dict | reactions
dict | timeline_url
stringlengths 61
62
| performed_via_github_app
null | state_reason
stringclasses 3
values | is_pull_request
bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/NVIDIA/NeMo/issues/2327 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2327/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2327/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2327/events | https://github.com/NVIDIA/NeMo/issues/2327 | 916,328,884 | MDU6SXNzdWU5MTYzMjg4ODQ= | 2,327 | Exporting PunctuationCapitalizationModel model to onnx | {
"login": "marlon-br",
"id": 12508098,
"node_id": "MDQ6VXNlcjEyNTA4MDk4",
"avatar_url": "https://avatars.githubusercontent.com/u/12508098?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/marlon-br",
"html_url": "https://github.com/marlon-br",
"followers_url": "https://api.github.com/users/marlon-br/followers",
"following_url": "https://api.github.com/users/marlon-br/following{/other_user}",
"gists_url": "https://api.github.com/users/marlon-br/gists{/gist_id}",
"starred_url": "https://api.github.com/users/marlon-br/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/marlon-br/subscriptions",
"organizations_url": "https://api.github.com/users/marlon-br/orgs",
"repos_url": "https://api.github.com/users/marlon-br/repos",
"events_url": "https://api.github.com/users/marlon-br/events{/privacy}",
"received_events_url": "https://api.github.com/users/marlon-br/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"And if I change `punctuation_en_distilbert.nemo` to `punctuation_en_bert.nemo` than .onnx model has 3 inputs, but the issue with \r\n`onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Unexpected input data type. Actual: (tensor(float)) , expected: (tensor(int64))` for `\"attention_mask\"` still exist",
"In forward DistilBERT takes only input_ids and attention_mask (no token_type_ids as in BERT).\r\n@borisfom could you help with the ONNX part?",
"Well, finally I ended up with the next code:\r\n\r\nExport .onnx and config yaml:\r\n\r\n```\r\nmodel = PunctuationCapitalizationModel.restore_from(restore_path=\"punctuation_en_distilbert.nemo\")\r\nmodel.export(\"punctuation_en_distilbert.onnx\")\r\n\r\nmodel = PunctuationCapitalizationModel.restore_from(restore_path=\"punctuation_en_distilbert.nemo\", return_config = True)\r\ntextfile = open(\"punctuation_en_distilbert.yaml\", \"w\")\r\ntextfile.write(str(OmegaConf.to_yaml(model)))\r\ntextfile.close()\r\n```\r\n\r\nExport tokenizer:\r\n```\r\nfrom transformers import BertTokenizer\r\ntokenizer = BertTokenizer.from_pretrained('distilbert-base-uncased')\r\ntokenizer.save_pretrained(\"./my_model_directory/\")\r\n```\r\nand rename `tokenizer_config.json` to `distilbert-base-uncased.json`\r\nAlso need to download https://huggingface.co/distilbert-base-uncased/blob/main/tokenizer.json to the my_model_directory folder and change `distilbert-base-uncased.json` content a bit, set `\"tokenizer_file\"` to `\"tokenizer.json\"`\r\nNow make processing with onnx:\r\n\r\n```\r\nimport numpy as np\r\nimport torch\r\nimport yaml\r\nfrom nemo.collections.common.tokenizers import AutoTokenizer\r\nfrom nemo.collections.nlp.data.token_classification.punctuation_capitalization_dataset import \\\r\n BertPunctuationCapitalizationInferDataset\r\nimport onnxruntime\r\n\r\nimport logging\r\nlogging.getLogger('nemo_logger').setLevel(logging.ERROR)\r\n\r\ndef create_infer_dataloader(tokenizer, queries):\r\n batch_size = len(queries)\r\n\r\n dataset = BertPunctuationCapitalizationInferDataset(\r\n tokenizer=tokenizer, queries=queries, max_seq_length=512\r\n )\r\n return torch.utils.data.DataLoader(\r\n dataset=dataset,\r\n collate_fn=dataset.collate_fn,\r\n batch_size=batch_size,\r\n shuffle=False,\r\n num_workers=1,\r\n drop_last=False,\r\n )\r\n\r\ndef to_numpy(tensor):\r\n return tensor.detach().cpu().numpy() if tensor.requires_grad else tensor.cpu().numpy()\r\n\r\nqueries = [\"by the end of no such thing the audience like beatrice has a watchful affection for the monster\", \\\r\n \"let me guess you're the kind of guy that ignores the rules cause it makes you feel in control am i right \\\r\n you're not wrong you think that's cute do you think it's cute\"]\r\n\r\ntokenizer = AutoTokenizer(pretrained_model_name=\"./my_model_directory/distilbert-base-uncased.json\", \\\r\n vocab_file=\"./my_model_directory/vocab.txt\")\r\n\r\ndata_loader = create_infer_dataloader(tokenizer, queries)\r\n\r\nwith open(\"punctuation_en_distilbert.yaml\") as f:\r\n params = yaml.safe_load(f)\r\n\r\nall_punct_preds = []\r\nall_capit_preds = []\r\n\r\noptions = onnxruntime.SessionOptions()\r\noptions.intra_op_num_threads = 1\r\n\r\nsess_punct = onnxruntime.InferenceSession(\"punctuation_en_distilbert.onnx\", options, providers=[\"CPUExecutionProvider\"]) \r\n\r\n# most of processing was taken from \r\n# https://github.com/NVIDIA/NeMo/blob/5839aee402f314aa413b28e9042b1e1cac10a114/nemo/collections/nlp/models/token_classification/punctuation_capitalization_model.py#L403\r\nfor batch in data_loader:\r\n input_ids, input_type_ids, input_mask, subtokens_mask = batch\r\n input_mask = input_mask.type(torch.int64) # why?\r\n ort_inputs = {\"attention_mask\": to_numpy(input_mask), \"input_ids\": to_numpy(input_ids)}\r\n\r\n punct_logits, capit_logits = sess_punct.run(None, input_feed = ort_inputs)\r\n\r\n subtokens_mask = subtokens_mask > 0.5\r\n\r\n punct_preds = [\r\n list(p_l[subtokens_mask[i]]) for i, p_l in enumerate(np.argmax(punct_logits, axis=-1))\r\n ]\r\n capit_preds = [\r\n list(c_l[subtokens_mask[i]]) for i, c_l in enumerate(np.argmax(capit_logits, axis=-1))\r\n ]\r\n\r\n all_punct_preds.extend(punct_preds)\r\n all_capit_preds.extend(capit_preds)\r\n\r\npunct_ids_to_labels = {v: k for k, v in params['punct_label_ids'].items()}\r\ncapit_ids_to_labels = {v: k for k, v in params['capit_label_ids'].items()}\r\n\r\nqueries = [q.strip().split() for q in queries]\r\n\r\nresult = []\r\n\r\nfor i, query in enumerate(queries):\r\n punct_preds = all_punct_preds[i]\r\n capit_preds = all_capit_preds[i]\r\n if len(query) != len(punct_preds):\r\n # removing the end of phrase punctuation of the truncated segment\r\n punct_preds[-1] = 0\r\n max_len = len(punct_preds)\r\n query = query[:max_len]\r\n\r\n query_with_punct_and_capit = ''\r\n for j, word in enumerate(query):\r\n punct_label = punct_ids_to_labels[punct_preds[j]]\r\n capit_label = capit_ids_to_labels[capit_preds[j]]\r\n\r\n if capit_label != params['dataset']['pad_label']:\r\n word = word.capitalize()\r\n query_with_punct_and_capit += word\r\n if punct_label != params['dataset']['pad_label']:\r\n query_with_punct_and_capit += punct_label\r\n query_with_punct_and_capit += ' '\r\n\r\n result.append(query_with_punct_and_capit.strip())\r\n```\r\n\r\nnot sure if I did everything correct with tokenizer export and import, but it works ",
"I am still facing the same issue after following the steps. ValueError: Model requires 3 inputs. Input Feed contains 2 \r\nDid I miss something?"
] | 2021-06-09T15:17:36 | 2021-12-16T13:09:13 | 2021-06-10T10:38:31 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I am trying to export the nemo_nlp PunctuationCapitalizationModel to onnx format.
I was able to do so using this block of code:
```
model = PunctuationCapitalizationModel.restore_from(restore_path="punctuation_en_distilbert.nemo")
model.export("punctuation_en_distilbert.onnx")
```
This works fine and I get the onnx file and I am able to load it using onnx.load and check it using onnx.checker.check_model.
The issues are that exported onnx model has only 2 inputs:
1.
```
[name: "input_ids"
type {
tensor_type {
elem_type: 7
shape {
dim {
dim_param: "input_ids_dynamic_axes_1"
}
dim {
dim_param: "input_ids_dynamic_axes_2"
}
}
}
}
, name: "attention_mask"
type {
tensor_type {
elem_type: 7
shape {
dim {
dim_param: "attention_mask_dynamic_axes_1"
}
dim {
dim_param: "attention_mask_dynamic_axes_2"
}
}
}
}
]
```
and
2. if I call it like:
```
options = onnxruntime.SessionOptions()
options.intra_op_num_threads = 1
sess_punct = onnxruntime.InferenceSession("punctuation_en_distilbert.onnx", options, providers=["CPUExecutionProvider"])
modelPunct = PunctuationCapitalizationModel.restore_from("punctuation_en_distilbert.nemo")
data_loader = create_infer_dataloader(modelPunct.tokenizer, queries)
for batch in data_loader:
input_ids, input_type_ids, input_mask, subtokens_mask = batch
ort_inputs = {"input_ids": to_numpy(input_ids),"attention_mask":to_numpy(input_mask)}
punct_logits, capit_logits = sess_punct.run(None, input_feed = ort_inputs)
```
where `create_infer_dataloader()` is
```
def create_infer_dataloader(tokenizer, queries):
batch_size = len(queries)
dataset = BertPunctuationCapitalizationInferDataset(
tokenizer=tokenizer, queries=queries, max_seq_length=512
)
return torch.utils.data.DataLoader(
dataset=dataset,
collate_fn=dataset.collate_fn,
batch_size=batch_size,
shuffle=False,
num_workers=1,
drop_last=False,
)
```
it gives me an error: `onnxruntime.capi.onnxruntime_pybind11_state.InvalidArgument: [ONNXRuntimeError] : 2 : INVALID_ARGUMENT : Unexpected input data type. Actual: (tensor(float)) , expected: (tensor(int64))`
So the question is what is wrong with the exported model?
Why the first question is why there are only 2 input parameters instead of 3, that we can see in different tutorials or source code of the PunctuationCapitalizationModel model and why data types are messed | {
"login": "marlon-br",
"id": 12508098,
"node_id": "MDQ6VXNlcjEyNTA4MDk4",
"avatar_url": "https://avatars.githubusercontent.com/u/12508098?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/marlon-br",
"html_url": "https://github.com/marlon-br",
"followers_url": "https://api.github.com/users/marlon-br/followers",
"following_url": "https://api.github.com/users/marlon-br/following{/other_user}",
"gists_url": "https://api.github.com/users/marlon-br/gists{/gist_id}",
"starred_url": "https://api.github.com/users/marlon-br/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/marlon-br/subscriptions",
"organizations_url": "https://api.github.com/users/marlon-br/orgs",
"repos_url": "https://api.github.com/users/marlon-br/repos",
"events_url": "https://api.github.com/users/marlon-br/events{/privacy}",
"received_events_url": "https://api.github.com/users/marlon-br/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2327/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/2327/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2326 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2326/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2326/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2326/events | https://github.com/NVIDIA/NeMo/pull/2326 | 915,885,124 | MDExOlB1bGxSZXF1ZXN0NjY1NjQxMTM5 | 2,326 | Add notebook with recommendations for 8 kHz speech | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-09T07:28:30 | 2022-01-11T16:37:10 | 2021-06-18T03:00:42 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2326",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2326",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2326.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2326.patch",
"merged_at": "2021-06-18T03:00:42"
} | {
"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/2326/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/2326/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2325 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2325/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2325/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2325/events | https://github.com/NVIDIA/NeMo/issues/2325 | 915,798,666 | MDU6SXNzdWU5MTU3OTg2NjY= | 2,325 | How to use asr model for streaming | {
"login": "ravi02512",
"id": 46443549,
"node_id": "MDQ6VXNlcjQ2NDQzNTQ5",
"avatar_url": "https://avatars.githubusercontent.com/u/46443549?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ravi02512",
"html_url": "https://github.com/ravi02512",
"followers_url": "https://api.github.com/users/ravi02512/followers",
"following_url": "https://api.github.com/users/ravi02512/following{/other_user}",
"gists_url": "https://api.github.com/users/ravi02512/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ravi02512/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ravi02512/subscriptions",
"organizations_url": "https://api.github.com/users/ravi02512/orgs",
"repos_url": "https://api.github.com/users/ravi02512/repos",
"events_url": "https://api.github.com/users/ravi02512/events{/privacy}",
"received_events_url": "https://api.github.com/users/ravi02512/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | true | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"How to do real time streaming using Citrinet asr model? Will it be the same way as of Quartznet?",
"> How to do real time streaming using Citrinet asr model? Will it be the same way as of Quartznet?\r\n\r\ndo you know how to implement quartznet for real time streaming??",
"\r\n\r\n> > How to do real time streaming using Citrinet asr model? Will it be the same way as of Quartznet?\r\n> \r\n> do you know how to implement quartznet for real time streaming??\r\n\r\nNo, I didn't try yet by myself.",
"For streaming you need FrameAsr like from this notebook https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/02_Online_ASR_Microphone_Demo.ipynb or use Nvidia Jarvis",
"> For streaming you need FrameAsr like from this notebook https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/02_Online_ASR_Microphone_Demo.ipynb or use Nvidia Jarvis\r\n\r\nYes, I tried this earlier but it didn't work for CitriNet model.",
"What you mean by didn't work? Did you use correct stats? https://github.com/NVIDIA/NeMo/issues/1203\r\nAnd from what I know CitriNet use different stride,you might need to take that into account, I don't have experience with CitriNet yet",
"> What you mean by didn't work? Did you use correct stats? #1203\r\n> And from what I know CitriNet use different stride,you might need to take that into account, I don't have experience with CitriNet yet\r\n\r\nCorrect, It is working for Quartznet model, but no output was coming for Citrinet model. I wanted to know what could be the reason of it. @khursani8 thanx for your response, I'll be checking for different stride."
] | 2021-06-09T05:40:41 | 2021-06-09T23:22:50 | 2021-06-09T23:22:49 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello folks
I used nvidia quartznet for asr and it is taking wav files to transcribe. How to do real time streaming asr model
Thanks | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2325/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/2325/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2324 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2324/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2324/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2324/events | https://github.com/NVIDIA/NeMo/pull/2324 | 915,656,354 | MDExOlB1bGxSZXF1ZXN0NjY1NDQ5MDg3 | 2,324 | change tensor interface to B,T,D between mel generator and vocoder | {
"login": "junkin",
"id": 128148,
"node_id": "MDQ6VXNlcjEyODE0OA==",
"avatar_url": "https://avatars.githubusercontent.com/u/128148?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/junkin",
"html_url": "https://github.com/junkin",
"followers_url": "https://api.github.com/users/junkin/followers",
"following_url": "https://api.github.com/users/junkin/following{/other_user}",
"gists_url": "https://api.github.com/users/junkin/gists{/gist_id}",
"starred_url": "https://api.github.com/users/junkin/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/junkin/subscriptions",
"organizations_url": "https://api.github.com/users/junkin/orgs",
"repos_url": "https://api.github.com/users/junkin/repos",
"events_url": "https://api.github.com/users/junkin/events{/privacy}",
"received_events_url": "https://api.github.com/users/junkin/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Please hold off merge of this until we make sure neural type information is included in nemo files.",
"> Please hold off merge of this until we make sure neural type information is included in nemo files.\r\n\r\n@ryanleary @junkin Should we try to merge this into NeMo? Or do you still want to block for now?",
"@blisc @junkin and @ryanleary what's the status of this? Do we need this PR? If not, someone, please close"
] | 2021-06-09T00:12:12 | 2022-09-22T18:46:32 | 2022-09-22T18:46:32 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2324",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2324",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2324.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2324.patch",
"merged_at": null
} | updates for TTS models to generate mels with output of type B, T, D and all vocoders to take as input B,T,D shapes to enable continues memory when grabbing chunks of spectrogram at inference time. | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2324/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/2324/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2323 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2323/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2323/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2323/events | https://github.com/NVIDIA/NeMo/pull/2323 | 915,610,283 | MDExOlB1bGxSZXF1ZXN0NjY1NDA5Mzcw | 2,323 | sgdqa update data directories for testing | {
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging 265bf1cf644ca545c07361b0c69b139176a04c16 into cfe2548d4db9dcc4ca393ee63604139c9f681cd8 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-daa3ed728ff039b73f402e72c8d2b2aa4efa1369)\n\n**new alerts:**\n\n* 2 for Unused exception object"
] | 2021-06-08T22:48:30 | 2021-06-16T18:23:59 | 2021-06-16T18:23:55 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2323",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2323",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2323.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2323.patch",
"merged_at": "2021-06-16T18:23:55"
} | - [x ] is backwards compatible
- [x ] fixes loading previously trained .nemo model for testing
Signed-off-by: Yang Zhang <[email protected]> | {
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2323/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/2323/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2322 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2322/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2322/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2322/events | https://github.com/NVIDIA/NeMo/issues/2322 | 915,472,742 | MDU6SXNzdWU5MTU0NzI3NDI= | 2,322 | how to use language model in train | {
"login": "Iagodel",
"id": 47003942,
"node_id": "MDQ6VXNlcjQ3MDAzOTQy",
"avatar_url": "https://avatars.githubusercontent.com/u/47003942?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Iagodel",
"html_url": "https://github.com/Iagodel",
"followers_url": "https://api.github.com/users/Iagodel/followers",
"following_url": "https://api.github.com/users/Iagodel/following{/other_user}",
"gists_url": "https://api.github.com/users/Iagodel/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Iagodel/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Iagodel/subscriptions",
"organizations_url": "https://api.github.com/users/Iagodel/orgs",
"repos_url": "https://api.github.com/users/Iagodel/repos",
"events_url": "https://api.github.com/users/Iagodel/events{/privacy}",
"received_events_url": "https://api.github.com/users/Iagodel/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"what models are you training?",
"We are using own dataset, with audios (wav) and .txt, in Portuguese. Like a small LibriSpeech",
"Language model is trained separately from the acoustic model such as QuartzNet. Then you can use LM during beam search process to improve the results. Please refer to this doc https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/asr_language_modeling.html ",
"Thanks @okuchaiev, this help us a lot and confirm something that we was thinking. \r\nNow we tried to use the same example in this doc and we got that:\r\n\r\n\r\nWe use beam search with lm:\r\nbeam_search_with_lm = nemo_asr.BeamSearchDecoderWithLM(\r\n vocab=labels,\r\n beam_width=beam_width,\r\n alpha=alpha,\r\n beta=beta,\r\n lm_path=lm_path,\r\n num_cpus=max(os.cpu_count(),1))\r\n \r\nI don't know if have anything wrong",
"The documentation above links to a utility script that will perform beam search for you - please try with that. The KenLM binary built for character models via that script should be evaluated using that script only. ",
"Thank you all for help, we got something news and got some results.\r\nNow, I just want one more help. About the parameters, beam_width, alpha and beta. Do have a documentation about that, blog or something that I can study about, to understand more how it work?",
"Beam width = beam size. Larger beam size will take longer for marginally better results. \r\nAlpha = LM weight. \r\nBeta = Length penalty. \r\nWe perform a sparse grid search over alpha and beta, keeping beam width fixed. "
] | 2021-06-08T20:37:28 | 2021-06-19T04:34:05 | 2021-06-19T04:34:05 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | We are trying to train models with a language model '.arpa', but reading the documentation and trying some scripts we can't get anything but the same result without the language model. I think we are doing something wrong. Do you have any tip for this question, another documentation or something like this to help us? Do we just use at the inference or not?
Regards | {
"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/2322/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/2322/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2321 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2321/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2321/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2321/events | https://github.com/NVIDIA/NeMo/issues/2321 | 915,382,319 | MDU6SXNzdWU5MTUzODIzMTk= | 2,321 | timestamp and score per word or sentence | {
"login": "AllanBraga14",
"id": 85580379,
"node_id": "MDQ6VXNlcjg1NTgwMzc5",
"avatar_url": "https://avatars.githubusercontent.com/u/85580379?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AllanBraga14",
"html_url": "https://github.com/AllanBraga14",
"followers_url": "https://api.github.com/users/AllanBraga14/followers",
"following_url": "https://api.github.com/users/AllanBraga14/following{/other_user}",
"gists_url": "https://api.github.com/users/AllanBraga14/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AllanBraga14/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AllanBraga14/subscriptions",
"organizations_url": "https://api.github.com/users/AllanBraga14/orgs",
"repos_url": "https://api.github.com/users/AllanBraga14/repos",
"events_url": "https://api.github.com/users/AllanBraga14/events{/privacy}",
"received_events_url": "https://api.github.com/users/AllanBraga14/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"We don't (yet) have something that sophisticated as an API that emits word level confidence. However, we do have CTC alignments available. You can use the CTC model.transcribe(..., return_hypotheses=True) argument, which will return a hypothesis object with a value in it called `alignment` - that describes what token the model predicted at each timestep of the Acoustic sequence. Note that this will include the CTC blank token (which is by conversion the last token in the decoding vocabulary).\r\n\r\nYou also get the log probabilities, so I suppose you could perform softmax at each timestep to compute the actual confidence per token predicted. \r\n\r\nStart and end times can be estimated as well depending on token boundaries and reversing the timesteps count by number of stride steps the model performs. Note that QuartzNet is a 2x stride model where as Citrinet is an 8x stride model. ",
"@titu1994 \r\nIs there any way to calculate the word timestep after applying beamsearch decoding on top of EncDecCTCModelBPE citrinet_1024.\r\n\r\nI tried [parlance ](https://github.com/parlance/ctcdecode) solution but took a very long time to execute, maybe because of the large vocabulary size."
] | 2021-06-08T19:19:49 | 2021-07-08T13:25:56 | 2021-06-09T21:25:18 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | We are doing training/transcription tests using NeMo for the Portuguese language and we are getting great results. But now I want to evolve the transcripts in order to get the confidence rate, timestamp of each word or sentence. Is there any tool inside NeMo that already does this autonomously?
Just for comparison, other transcribers return Json similar to this:
"words": [
{
"end_time": 0.39077207,
"score": 66,
"start_time": 0.272892,
"text": "a"
},
{
"end_time": 1.1700554,
"score": 94,
"start_time": 0.40797436,
"text": "internet"
}
] | {
"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/2321/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/2321/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2320 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2320/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2320/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2320/events | https://github.com/NVIDIA/NeMo/pull/2320 | 915,267,091 | MDExOlB1bGxSZXF1ZXN0NjY1MTAwOTk1 | 2,320 | sparrowhawk tests + punctuation post processing for pynini TN | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-08T17:12:20 | 2021-06-11T03:58:26 | 2021-06-11T03:58:23 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2320",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2320",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2320.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2320.patch",
"merged_at": "2021-06-11T03:58:23"
} | {
"login": "ekmb",
"id": 10428420,
"node_id": "MDQ6VXNlcjEwNDI4NDIw",
"avatar_url": "https://avatars.githubusercontent.com/u/10428420?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ekmb",
"html_url": "https://github.com/ekmb",
"followers_url": "https://api.github.com/users/ekmb/followers",
"following_url": "https://api.github.com/users/ekmb/following{/other_user}",
"gists_url": "https://api.github.com/users/ekmb/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ekmb/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ekmb/subscriptions",
"organizations_url": "https://api.github.com/users/ekmb/orgs",
"repos_url": "https://api.github.com/users/ekmb/repos",
"events_url": "https://api.github.com/users/ekmb/events{/privacy}",
"received_events_url": "https://api.github.com/users/ekmb/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2320/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/2320/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2319 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2319/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2319/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2319/events | https://github.com/NVIDIA/NeMo/pull/2319 | 915,225,796 | MDExOlB1bGxSZXF1ZXN0NjY1MDY2MDI5 | 2,319 | [BUGFIX] OmegaConf forward compatibility | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-08T16:19:59 | 2021-06-08T17:40:43 | 2021-06-08T17:40:35 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2319",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2319",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2319.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2319.patch",
"merged_at": "2021-06-08T17:40:35"
} | Apply updates where OmegaConf is breaking with 2.1.0.
See #2316 for the original PR to v1.0.0.
Note, we have to upper bound OmegaConf as Hydra 1.0.6 is not compatible with OmegaConf 2.1.0. When the new version of Hydra is released we should remove the upper bound. | {
"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/2319/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/2319/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2318 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2318/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2318/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2318/events | https://github.com/NVIDIA/NeMo/pull/2318 | 915,087,463 | MDExOlB1bGxSZXF1ZXN0NjY0OTQ3MDg1 | 2,318 | Exportable fastpitch | {
"login": "ryanleary",
"id": 2212584,
"node_id": "MDQ6VXNlcjIyMTI1ODQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/2212584?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ryanleary",
"html_url": "https://github.com/ryanleary",
"followers_url": "https://api.github.com/users/ryanleary/followers",
"following_url": "https://api.github.com/users/ryanleary/following{/other_user}",
"gists_url": "https://api.github.com/users/ryanleary/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ryanleary/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ryanleary/subscriptions",
"organizations_url": "https://api.github.com/users/ryanleary/orgs",
"repos_url": "https://api.github.com/users/ryanleary/repos",
"events_url": "https://api.github.com/users/ryanleary/events{/privacy}",
"received_events_url": "https://api.github.com/users/ryanleary/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@blisc @borisfom Think we can get this merged?",
"> @blisc @borisfom Think we can get this merged?\r\n\r\nCan you merge with main?",
"@blisc merged in the latest changes from main, so this should have the remaining deltas only now."
] | 2021-06-08T14:01:38 | 2021-06-14T14:00:55 | 2021-06-14T14:00:55 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2318",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2318",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2318.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2318.patch",
"merged_at": null
} | {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2318/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/2318/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2317 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2317/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2317/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2317/events | https://github.com/NVIDIA/NeMo/issues/2317 | 914,406,197 | MDU6SXNzdWU5MTQ0MDYxOTc= | 2,317 | conv asr onnx export no signal length input | {
"login": "Slyne",
"id": 6286804,
"node_id": "MDQ6VXNlcjYyODY4MDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/6286804?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Slyne",
"html_url": "https://github.com/Slyne",
"followers_url": "https://api.github.com/users/Slyne/followers",
"following_url": "https://api.github.com/users/Slyne/following{/other_user}",
"gists_url": "https://api.github.com/users/Slyne/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Slyne/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Slyne/subscriptions",
"organizations_url": "https://api.github.com/users/Slyne/orgs",
"repos_url": "https://api.github.com/users/Slyne/repos",
"events_url": "https://api.github.com/users/Slyne/events{/privacy}",
"received_events_url": "https://api.github.com/users/Slyne/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": "borisfom",
"id": 14189615,
"node_id": "MDQ6VXNlcjE0MTg5NjE1",
"avatar_url": "https://avatars.githubusercontent.com/u/14189615?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/borisfom",
"html_url": "https://github.com/borisfom",
"followers_url": "https://api.github.com/users/borisfom/followers",
"following_url": "https://api.github.com/users/borisfom/following{/other_user}",
"gists_url": "https://api.github.com/users/borisfom/gists{/gist_id}",
"starred_url": "https://api.github.com/users/borisfom/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/borisfom/subscriptions",
"organizations_url": "https://api.github.com/users/borisfom/orgs",
"repos_url": "https://api.github.com/users/borisfom/repos",
"events_url": "https://api.github.com/users/borisfom/events{/privacy}",
"received_events_url": "https://api.github.com/users/borisfom/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "borisfom",
"id": 14189615,
"node_id": "MDQ6VXNlcjE0MTg5NjE1",
"avatar_url": "https://avatars.githubusercontent.com/u/14189615?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/borisfom",
"html_url": "https://github.com/borisfom",
"followers_url": "https://api.github.com/users/borisfom/followers",
"following_url": "https://api.github.com/users/borisfom/following{/other_user}",
"gists_url": "https://api.github.com/users/borisfom/gists{/gist_id}",
"starred_url": "https://api.github.com/users/borisfom/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/borisfom/subscriptions",
"organizations_url": "https://api.github.com/users/borisfom/orgs",
"repos_url": "https://api.github.com/users/borisfom/repos",
"events_url": "https://api.github.com/users/borisfom/events{/privacy}",
"received_events_url": "https://api.github.com/users/borisfom/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"I'm not very familiar with ONNX, @borisfom could you take a look at this ?",
"@borisfom would comment on the onnx export and why length is ignored. But AFAIK with QuartzNet you do not need to pass the lengths as they are not used at all in the forward pass (unless you enable the masked convolutions). You just need to mask-out the outputs in the decoding procedure. For Conformer, lengths are necessary as we need to mask the paddings in self-attention.",
"@VahidooX I checked the [config file](https://github.com/NVIDIA/NeMo/blob/main/examples/asr/conf/quartznet/quartznet_15x5.yaml) for Quartznet and it uses `conv_mask: True`. It would be better to give users an option to decide whether to have input length or not.",
"@borisfom Any update ?",
"@titu1994 Do you mind I add seq length to \r\n\r\nhttps://github.com/NVIDIA/NeMo/blob/7ef1782b94386629fbfafece72f618096c33a9f3/nemo/collections/asr/modules/conv_asr.py#L259",
"I don't know the impact this will have on downstream tools that require the onnx file in current format.\r\n\r\n@borisfom please advise here ",
" @Slyne : the change needs to be more extensive QuartzNet removes masked convolutions for inference, and you would need to keep them, too : \r\n def _prepare_for_export(self, **kwargs):\r\n m_count = 0\r\n for m in self.modules():\r\n if isinstance(m, MaskedConv1d):\r\n m.use_mask = False\r\n m_count += 1\r\n Exportable._prepare_for_export(self, **kwargs)\r\n logging.warning(f\"Turned off {m_count} masked convolutions\")\r\n\r\nThis has to be controlled by some other flag, not the main conv_mask - because there are currently many nets that do have this flag as True for training but do not need that for inference. \r\nFor your local experiments, you can try hacking it out - please also note we originally had to use removal of masked convolutions to get the code exported in ONNX at all - that may have been fixed, but I did not check.",
"@borisfom \r\n\"This has to be controlled by some other flag, not the main conv_mask - because there are currently many nets that do have this flag as True for training but do not need that for inference.\"\r\n\r\nDo you mean the sequence length will not affect the inference procedure for quartznet ? Or we can just trim the quartznet output by sequence length?"
] | 2021-06-08T06:14:51 | 2021-10-26T23:45:33 | 2021-10-26T23:45:33 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
I'm trying to export Quartznet model to onnx by using asr_model.export() function. However, the generated onnx model doesn't have the input length as an input and it only has 'audio_signal'. Therefore after converting the onnx model to trt engine, I find the inference results have a WER much worse than inferencing using NeMo directly and the trt engine is vulnerable to different batch sizes. Could you help confirm whether this is a bug ?
I find that in conformer encoder, [the input_example](https://github.com/NVIDIA/NeMo/blob/7ef1782b94386629fbfafece72f618096c33a9f3/nemo/collections/asr/modules/conformer_encoder.py#L75) used to generate onnx model has input example length. However, in [conv asr](https://github.com/NVIDIA/NeMo/blob/7ef1782b94386629fbfafece72f618096c33a9f3/nemo/collections/asr/modules/conv_asr.py#L250), there's only input example.
**Steps/Code to reproduce bug**
```bash
# generate calibrated model
python3 speech_to_text_calibrate.py --dataset=/raid/data/LibriSpeech/manifest.dev-other --asr_model=QuartzNet15x5Base-En
# generate onnx model
python3 speech_to_text_quant_infer.py --dataset=/raid/data/LibriSpeech/manifest.dev-other --asr_model=QuartzNet15x5Base-En-max-256.nemo --onnx
# Got WER 10.7x%
python3 speech_to_text_quant_infer_trt.py --dataset=/raid/data/LibriSpeech/manifest.dev-other --asr_model=QuartzNet15x5Base-En-max-256.nemo --asr_onnx=./QuartzNet15x5Base-En-max-256.onnx --qat
# Got WER 11.x%
```
**Expected behavior**
Expect the `speech_to_text_quant_infer.py` to have the same result as `speech_to_text_quant_infer_trt.py`
**Environment overview (please complete the following information)**
nvcr.io/nvidia/pytorch:21.05-py3
+ Tensorrt 8.0.0.3
+ NeMo main
| {
"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/2317/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/2317/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2316 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2316/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2316/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2316/events | https://github.com/NVIDIA/NeMo/pull/2316 | 913,886,795 | MDExOlB1bGxSZXF1ZXN0NjYzODU1NDY2 | 2,316 | Pin OmegaConf version for 1.0.0 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-07T19:26:02 | 2021-06-08T23:40:58 | 2021-06-08T00:22:15 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2316",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2316",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2316.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2316.patch",
"merged_at": "2021-06-08T00:22:15"
} | # Bugfix
- Update OmageConf signature to correctly save and restore models in OmegaConf 2.1.0 release
Signed-off-by: smajumdar <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2316/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/2316/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2315 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2315/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2315/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2315/events | https://github.com/NVIDIA/NeMo/pull/2315 | 913,730,138 | MDExOlB1bGxSZXF1ZXN0NjYzNzE4MzQ2 | 2,315 | Shallow fusion | {
"login": "AlexGrinch",
"id": 8689095,
"node_id": "MDQ6VXNlcjg2ODkwOTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/8689095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AlexGrinch",
"html_url": "https://github.com/AlexGrinch",
"followers_url": "https://api.github.com/users/AlexGrinch/followers",
"following_url": "https://api.github.com/users/AlexGrinch/following{/other_user}",
"gists_url": "https://api.github.com/users/AlexGrinch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AlexGrinch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AlexGrinch/subscriptions",
"organizations_url": "https://api.github.com/users/AlexGrinch/orgs",
"repos_url": "https://api.github.com/users/AlexGrinch/repos",
"events_url": "https://api.github.com/users/AlexGrinch/events{/privacy}",
"received_events_url": "https://api.github.com/users/AlexGrinch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-07T16:27:43 | 2021-06-10T18:28:57 | 2021-06-10T18:28:57 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2315",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2315",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2315.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2315.patch",
"merged_at": "2021-06-10T18:28:57"
} | Shallow fusion with LM during NMT beam search generation.
1. Implements `BeamSearchSequenceGeneratorWithLanguageModel` for shallow fusion with external LM based on the same tokenizer as NMT model.
2. Updates `examples/nlp/machine_translation/nmt_transformer_infer.py` script to support LM shallow fusion. | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2315/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/2315/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2314 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2314/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2314/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2314/events | https://github.com/NVIDIA/NeMo/pull/2314 | 913,632,108 | MDExOlB1bGxSZXF1ZXN0NjYzNjM0MDIw | 2,314 | patch quantization | {
"login": "Slyne",
"id": 6286804,
"node_id": "MDQ6VXNlcjYyODY4MDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/6286804?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Slyne",
"html_url": "https://github.com/Slyne",
"followers_url": "https://api.github.com/users/Slyne/followers",
"following_url": "https://api.github.com/users/Slyne/following{/other_user}",
"gists_url": "https://api.github.com/users/Slyne/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Slyne/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Slyne/subscriptions",
"organizations_url": "https://api.github.com/users/Slyne/orgs",
"repos_url": "https://api.github.com/users/Slyne/repos",
"events_url": "https://api.github.com/users/Slyne/events{/privacy}",
"received_events_url": "https://api.github.com/users/Slyne/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 2 alerts** when merging f139082ed38998ea2474a09dc2f8759b68625969 into 29b6dd8890d5c196ded96351607ff6642406e75a - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-90934cd23206306438231b8b7dab39a60d08f70b)\n\n**new alerts:**\n\n* 2 for Unused import",
"@titu1994 Could you help check the above issue ?",
"I'm not particularly familiar with the quantization aware training methods. Will request someone from team to help with this. "
] | 2021-06-07T14:47:59 | 2021-06-08T02:53:32 | 2021-06-07T20:19:09 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2314",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2314",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2314.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2314.patch",
"merged_at": "2021-06-07T20:19:09"
} | Hi @titu1994
I've further tested the quantization part code and I get the below results:
First, calibrate the model:
```bash
python speech_to_text_calibrate.py --asr_model=./QuartzNet15x5-Zh-NR.nemo --dataset=/raid/data/data_aishell2/dev_manifest.json --dont_normalize_text
```
Then, test the quant model:
```bash
python3 speech_to_text_quant_infer.py --asr_model=./QuartzNet15x5-Zh-NR-max-256.nemo --dont_normalize_text --use_cer --onnx --dataset=/raid/data/data_aishell2/dev_manifest.json
```
I get a CER of 7.34%.
Finally, to test with trt with the generated onnx model:
```bash
python3 speech_to_text_quant_infer_trt.py --asr_model=./QuartzNet15x5-Zh-NR.nemo --asr_onnx=./QuartzNet15x5-Zh-NR-max-256.onnx --dont_normalize_text --use_cer --qat --dataset=/raid/data/data_aishell2/dev_manifest.json --batch_size=256
```
I get a CER of 9.08% with batch_size=256, 8.06% with batch_size=256 & sort dev set by audio length, 7.95% with batch_size=1.
I know it is normal to have 0.01%~0.05% error tolerance due to the padding of Quartznet. But the above result shows that there might be some issues with the script or the generated trt engine file.
(The size of the generated engine file is 24MB, which seems reasonable and the original model trt engine file is 95MB.)
I build my environment by starting from `nvcr.io/nvidia/pytorch:21.05-py3` + tensorrt 8 EA + nemo_toolkit[all].
Not sure if I made any mistakes in the above steps. Please correct me.
| {
"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/2314/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/2314/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2313 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2313/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2313/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2313/events | https://github.com/NVIDIA/NeMo/issues/2313 | 913,275,367 | MDU6SXNzdWU5MTMyNzUzNjc= | 2,313 | Changing Language Model in QuatzNet5*15 | {
"login": "rose768",
"id": 85157007,
"node_id": "MDQ6VXNlcjg1MTU3MDA3",
"avatar_url": "https://avatars.githubusercontent.com/u/85157007?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rose768",
"html_url": "https://github.com/rose768",
"followers_url": "https://api.github.com/users/rose768/followers",
"following_url": "https://api.github.com/users/rose768/following{/other_user}",
"gists_url": "https://api.github.com/users/rose768/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rose768/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rose768/subscriptions",
"organizations_url": "https://api.github.com/users/rose768/orgs",
"repos_url": "https://api.github.com/users/rose768/repos",
"events_url": "https://api.github.com/users/rose768/events{/privacy}",
"received_events_url": "https://api.github.com/users/rose768/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"BERT is a masked LM and different from regular Transformer LM. Using a maksed LM model to score a sentence is not as efficient as Transformer. You need to call it X times and X is equal to the number of tokens in a candidate while Transformer just needs one pass. For instance, if average of your candidates have 30 tokens, BERT is 30 times slower than Transformer. It makes BERT very expensive for rescoring while the improvement may not be that much significant. There are some papers which have tried to address this issue like this one:\r\n\r\nhttps://assets.amazon.science/cc/54/980cb7d74f93849b49b9d5c42466/masked-language-model-scoring.pdf\r\n\r\n"
] | 2021-06-07T08:24:35 | 2021-07-13T15:27:32 | 2021-07-13T15:27:32 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi
I use QuartzNet15-5 for fine-tuning. Based on this [paper](https://arxiv.org/pdf/1910.10261.pdf) I understand that QuartzNet15-5 uses 6-gram and T-XL language models. How can I change the language model in QuartNet15-5? For instance, using Bert.
**Environment overview**
Environment location: Google Colab
Method of NeMo install: !pip install nemo_toolkit[asr]
NeMo version: 1.0.0
Learning Rate: 1e-3
**Environment details**
OS version : "Ubuntu20.04.3 LTS"
PyTorch version : "1.7.1"
| {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2313/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/2313/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2312 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2312/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2312/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2312/events | https://github.com/NVIDIA/NeMo/pull/2312 | 912,303,048 | MDExOlB1bGxSZXF1ZXN0NjYyNDc2MTA2 | 2,312 | ddp translate GPU allocation fix | {
"login": "AlexGrinch",
"id": 8689095,
"node_id": "MDQ6VXNlcjg2ODkwOTU=",
"avatar_url": "https://avatars.githubusercontent.com/u/8689095?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/AlexGrinch",
"html_url": "https://github.com/AlexGrinch",
"followers_url": "https://api.github.com/users/AlexGrinch/followers",
"following_url": "https://api.github.com/users/AlexGrinch/following{/other_user}",
"gists_url": "https://api.github.com/users/AlexGrinch/gists{/gist_id}",
"starred_url": "https://api.github.com/users/AlexGrinch/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/AlexGrinch/subscriptions",
"organizations_url": "https://api.github.com/users/AlexGrinch/orgs",
"repos_url": "https://api.github.com/users/AlexGrinch/repos",
"events_url": "https://api.github.com/users/AlexGrinch/events{/privacy}",
"received_events_url": "https://api.github.com/users/AlexGrinch/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-05T15:05:45 | 2021-06-10T05:48:43 | 2021-06-10T05:48:43 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2312",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2312",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2312.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2312.patch",
"merged_at": "2021-06-10T05:48:43"
} | Replaced `model.to(rank)` with `torch.cuda.set_device(rank)` so several model instances do not stuck on 1st GPU. | {
"login": "MaximumEntropy",
"id": 9114321,
"node_id": "MDQ6VXNlcjkxMTQzMjE=",
"avatar_url": "https://avatars.githubusercontent.com/u/9114321?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MaximumEntropy",
"html_url": "https://github.com/MaximumEntropy",
"followers_url": "https://api.github.com/users/MaximumEntropy/followers",
"following_url": "https://api.github.com/users/MaximumEntropy/following{/other_user}",
"gists_url": "https://api.github.com/users/MaximumEntropy/gists{/gist_id}",
"starred_url": "https://api.github.com/users/MaximumEntropy/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/MaximumEntropy/subscriptions",
"organizations_url": "https://api.github.com/users/MaximumEntropy/orgs",
"repos_url": "https://api.github.com/users/MaximumEntropy/repos",
"events_url": "https://api.github.com/users/MaximumEntropy/events{/privacy}",
"received_events_url": "https://api.github.com/users/MaximumEntropy/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2312/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/2312/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2311 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2311/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2311/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2311/events | https://github.com/NVIDIA/NeMo/issues/2311 | 912,121,258 | MDU6SXNzdWU5MTIxMjEyNTg= | 2,311 | Is it possible to use BeamSearchDecoderWithLM() method in asr in a Windows 10 system? | {
"login": "computervisionpro",
"id": 40919247,
"node_id": "MDQ6VXNlcjQwOTE5MjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/40919247?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/computervisionpro",
"html_url": "https://github.com/computervisionpro",
"followers_url": "https://api.github.com/users/computervisionpro/followers",
"following_url": "https://api.github.com/users/computervisionpro/following{/other_user}",
"gists_url": "https://api.github.com/users/computervisionpro/gists{/gist_id}",
"starred_url": "https://api.github.com/users/computervisionpro/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/computervisionpro/subscriptions",
"organizations_url": "https://api.github.com/users/computervisionpro/orgs",
"repos_url": "https://api.github.com/users/computervisionpro/repos",
"events_url": "https://api.github.com/users/computervisionpro/events{/privacy}",
"received_events_url": "https://api.github.com/users/computervisionpro/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | {
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
},
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"We dont support Windows 10 in any capacity at the moment. ",
"I face the same problem in Colab, I installed all prerequisite."
] | 2021-06-05T06:40:36 | 2021-09-12T11:24:22 | 2021-07-13T15:27:22 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | I tried using the above feature in my Windows system, then I found that I need to execute the bash file in
`scripts/asr_language_modeling/ngram_lm/install_beamsearch_decoders.sh `
I also installed [swig for windows](http://www.swig.org/download.html), [wget for windows](https://eternallybored.org/misc/wget/) & [make for windows](https://sourceforge.net/projects/gnuwin32/files/make/3.81/)
additionally for running the wget & make commands in the .sh files.
After installing all these, I am getting an error while I am trying to run the above `install_beamsearch_decoders.sh ` inside the nemo folder.
`ModuleNotFoundError: No module named '_swig_decoders'`
Earlier, I was also getting error `ModuleNotFoundError: No module named 'swig_decoders'` but I resolved that by installing make & wget, so the command is able to generate swig_decoder.py file, but now I am getting `ModuleNotFoundError: No module named '_swig_decoders'` which is another python module file.
I have attached the log file with this. As per my understanding, this is happening mainly because in the log there comes an error
```
cl : Command line warning D9002 : ignoring unknown option '-O3'
cl : Command line warning D9002 : ignoring unknown option '-std=c++11'
```
[swig-error.txt](https://github.com/NVIDIA/NeMo/files/6602117/swig-error.txt)
However, I would appreciate if you can check the log file and see what is the problem if in case I am wrong. Also let me know if this is impossible to use in Windows system.
Thanks for the help! | {
"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/2311/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/2311/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2310 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2310/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2310/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2310/events | https://github.com/NVIDIA/NeMo/pull/2310 | 911,846,458 | MDExOlB1bGxSZXF1ZXN0NjYyMDYxMjk5 | 2,310 | get embedding for a single file | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-04T21:15:00 | 2021-06-05T20:05:41 | 2021-06-05T20:05:38 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2310",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2310",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2310.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2310.patch",
"merged_at": "2021-06-05T20:05:38"
} | Signed-off-by: nithinraok <[email protected]>
Usage:
`speaker_model = ExtractSpeakerEmbeddingsModel.from_pretrained(model_name="speakerverification_speakernet")`
` embs = speaker_model.get_embedding('audio_path')` | {
"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/2310/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/2310/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2309 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2309/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2309/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2309/events | https://github.com/NVIDIA/NeMo/pull/2309 | 911,761,719 | MDExOlB1bGxSZXF1ZXN0NjYxOTg5Mjc1 | 2,309 | Update container version to 21.05 | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-04T18:58:49 | 2021-06-14T23:40:59 | 2021-06-14T23:39:45 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2309",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2309",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2309.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2309.patch",
"merged_at": "2021-06-14T23:39:45"
} | # Changelog
- Update container version to 21.05 for CI and Dockerfile
Signed-off-by: smajumdar <[email protected]> | {
"login": "ericharper",
"id": 11999610,
"node_id": "MDQ6VXNlcjExOTk5NjEw",
"avatar_url": "https://avatars.githubusercontent.com/u/11999610?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ericharper",
"html_url": "https://github.com/ericharper",
"followers_url": "https://api.github.com/users/ericharper/followers",
"following_url": "https://api.github.com/users/ericharper/following{/other_user}",
"gists_url": "https://api.github.com/users/ericharper/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ericharper/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ericharper/subscriptions",
"organizations_url": "https://api.github.com/users/ericharper/orgs",
"repos_url": "https://api.github.com/users/ericharper/repos",
"events_url": "https://api.github.com/users/ericharper/events{/privacy}",
"received_events_url": "https://api.github.com/users/ericharper/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2309/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/2309/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2308 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2308/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2308/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2308/events | https://github.com/NVIDIA/NeMo/issues/2308 | 911,313,428 | MDU6SXNzdWU5MTEzMTM0Mjg= | 2,308 | Import Error on tutorial of Relation_Extraction-BioMegatron | {
"login": "blackbirt-5",
"id": 82014275,
"node_id": "MDQ6VXNlcjgyMDE0Mjc1",
"avatar_url": "https://avatars.githubusercontent.com/u/82014275?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blackbirt-5",
"html_url": "https://github.com/blackbirt-5",
"followers_url": "https://api.github.com/users/blackbirt-5/followers",
"following_url": "https://api.github.com/users/blackbirt-5/following{/other_user}",
"gists_url": "https://api.github.com/users/blackbirt-5/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blackbirt-5/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blackbirt-5/subscriptions",
"organizations_url": "https://api.github.com/users/blackbirt-5/orgs",
"repos_url": "https://api.github.com/users/blackbirt-5/repos",
"events_url": "https://api.github.com/users/blackbirt-5/events{/privacy}",
"received_events_url": "https://api.github.com/users/blackbirt-5/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 | [] | 2021-06-04T09:30:10 | 2021-10-26T23:45:17 | 2021-10-26T23:45:17 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Describe the bug**
Hi,
I have an error about tutorial of relation_extraction-Biomegatron.
In Colab, there was no problem to install and to test.
Step/Code : from nemo.collections import nlp as nemo_nlp
During import process, import error was happened like this.
ImportError: libhdf5.so.101: cannot open shared object file: No such file or directory
- Environment location: Linux centOS, Jupyter notebook.
- OS version: Linux CentOS
- PyTorch version: 1.7.1 (from nemo toolkit)
- Python version: 3.6.10
- GPU model (V100 on server)
Best Regards.
[ImportError.docx](https://github.com/NVIDIA/NeMo/files/6606953/ImportError.docx)
| {
"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/2308/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/2308/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2307 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2307/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2307/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2307/events | https://github.com/NVIDIA/NeMo/issues/2307 | 910,934,222 | MDU6SXNzdWU5MTA5MzQyMjI= | 2,307 | Is it possible to use ASR for online streaming with Beam Search? | {
"login": "Omarnabk",
"id": 72882909,
"node_id": "MDQ6VXNlcjcyODgyOTA5",
"avatar_url": "https://avatars.githubusercontent.com/u/72882909?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Omarnabk",
"html_url": "https://github.com/Omarnabk",
"followers_url": "https://api.github.com/users/Omarnabk/followers",
"following_url": "https://api.github.com/users/Omarnabk/following{/other_user}",
"gists_url": "https://api.github.com/users/Omarnabk/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Omarnabk/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Omarnabk/subscriptions",
"organizations_url": "https://api.github.com/users/Omarnabk/orgs",
"repos_url": "https://api.github.com/users/Omarnabk/repos",
"events_url": "https://api.github.com/users/Omarnabk/events{/privacy}",
"received_events_url": "https://api.github.com/users/Omarnabk/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | {
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "jbalam-nv",
"id": 4916480,
"node_id": "MDQ6VXNlcjQ5MTY0ODA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4916480?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jbalam-nv",
"html_url": "https://github.com/jbalam-nv",
"followers_url": "https://api.github.com/users/jbalam-nv/followers",
"following_url": "https://api.github.com/users/jbalam-nv/following{/other_user}",
"gists_url": "https://api.github.com/users/jbalam-nv/gists{/gist_id}",
"starred_url": "https://api.github.com/users/jbalam-nv/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/jbalam-nv/subscriptions",
"organizations_url": "https://api.github.com/users/jbalam-nv/orgs",
"repos_url": "https://api.github.com/users/jbalam-nv/repos",
"events_url": "https://api.github.com/users/jbalam-nv/events{/privacy}",
"received_events_url": "https://api.github.com/users/jbalam-nv/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
},
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"This is not supported. "
] | 2021-06-03T23:28:52 | 2021-08-17T01:03:42 | 2021-08-17T01:03:29 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | {
"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/2307/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/2307/timeline | null | completed | false |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2306 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2306/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2306/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2306/events | https://github.com/NVIDIA/NeMo/issues/2306 | 910,916,847 | MDU6SXNzdWU5MTA5MTY4NDc= | 2,306 | Export of rules to sparrowhawk | {
"login": "balacoon",
"id": 64325946,
"node_id": "MDEyOk9yZ2FuaXphdGlvbjY0MzI1OTQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/64325946?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/balacoon",
"html_url": "https://github.com/balacoon",
"followers_url": "https://api.github.com/users/balacoon/followers",
"following_url": "https://api.github.com/users/balacoon/following{/other_user}",
"gists_url": "https://api.github.com/users/balacoon/gists{/gist_id}",
"starred_url": "https://api.github.com/users/balacoon/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/balacoon/subscriptions",
"organizations_url": "https://api.github.com/users/balacoon/orgs",
"repos_url": "https://api.github.com/users/balacoon/repos",
"events_url": "https://api.github.com/users/balacoon/events{/privacy}",
"received_events_url": "https://api.github.com/users/balacoon/received_events",
"type": "Organization",
"user_view_type": "public",
"site_admin": false
} | [
{
"id": 1719393562,
"node_id": "MDU6TGFiZWwxNzE5MzkzNTYy",
"url": "https://api.github.com/repos/NVIDIA/NeMo/labels/feature",
"name": "feature",
"color": "68F3F0",
"default": false,
"description": "request/PR for a new feature"
}
] | closed | false | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"important fix - to export not normalizer.verbalizer.fst (which is VerbalizeFinalFst), but VerbalizeFst. There are a lot of other inconsistencies, but they seem to be related to serialization indeed:\r\n\r\nfailed to verbalize: \"time { hours: \"three\" minutes: \"thirty\" }\", because Unable to convert string to int32. I guess this is typing in proto.\r\n\r\nand another:\r\n\r\n[ERROR:protobuf_parser.cc:183] Unknown field: [integer_part]\r\n[ERROR:protobuf_parser.cc:342] Full input: [tokens { name: \"mister\" } tokens { name: \"Snookums\" } tokens { name: \"was\" } tokens { name: \"on\" } tokens { name: \"the\" } tokens { name: \"train\" } tokens { name: \"carrying\" } tokens { money { currency: \"dollars\" integer_part: \"forty\" fractional_part: \"two five\" } } tokens { name: \"(\" } tokens { money { currency: \"pounds\" integer_part: \"thirty\" fractional_part: \"six o\" } } tokens { name: \")\" } tokens { name: \"of\" } tokens { name: \"Belgian\" } tokens { name: \"chocolate\" } tokens { name: \"in\" } tokens { name: \"a\" } tokens { measure { cardinal { integer: \"three\" } units: \"kilograms\" } } tokens { name: \"box\" } tokens { name: \"that\" } tokens { name: \"was\" } tokens { measure { cardinal { integer: \"twenty\" } units: \"centimeters\" } } tokens { name: \"wide.\" }]\r\n[ERROR:normalizer.cc:131] Failed to parse tokens from FST for \"Mr. Snookums was on the train carrying $40.25 (£30.60) of Belgian chocolate in a 3kg box that was 20cm wide.\"",
"The documentation could be found [here](https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/v1.0.0/tools/text_processing_deployment.html) and tutorials: [Text normalization](https://colab.research.google.com/github/NVIDIA/NeMo/blob/main/tutorials/text_processing/Text_Normalization.ipynb) and [ITN](https://colab.research.google.com/github/NVIDIA/NeMo/blob/main/tutorials/text_processing/Inverse_Text_Normalization.ipynb)",
"@balacoon Did the instructions @ekmb helped you solve your problem?",
"yes, perfect. thanks!"
] | 2021-06-03T22:47:44 | 2021-06-16T00:57:54 | 2021-06-16T00:57:54 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | **Is your feature request related to a problem? Please describe.**
Paper for (inverse) text normalization describes that its easy to deploy developed text (de)normalization rules into prod using sparrowhawk. I havent found any documentation for that though. So far I figured how to export grammars as far files consumed by sparrowhawk:
```
normalizer = Normalizer(input_case="cased")
tag_exp = grm.Exporter(args.tagger)
tag_exp['TOKENIZE_AND_CLASSIFY'] = normalizer.tagger.fst
tag_exp.close()
verb_exp = grm.Exporter(args.verbalizer)
verb_exp['ALL'] = normalizer.verbalizer.fst
verb_exp.close()
```
But when I try to apply those rules, verbalization fails.
here is normalization using default sparrowhawk grammars shipped as part of documentation:
```
# cat test.txt
Hello 13 world
# normalizer_main --config=sparrowhawk_configuration.ascii_proto --multi_line_text < test.txt
[DEBUG:normalizer.cc:162] Verbalizing: ["13(r
13]
[DEBUG:normalizer.cc:195] Verbalize output: Words
Hello thirteen world
```
This is how it fails using grammars from NeMo:
```
[DEBUG:normalizer.cc:162] Verbalizing: ["13(r
thirteen]
[ERROR:rule_system.cc:121] Application of rule "ALL" failed[ERROR:normalizer.cc:213] Failed to verbalize "rthirteen"[WARNING:normalizer.cc:167] First-pass verbalization FAILED for ["13(rthirteen][ERROR:rule_system.cc:121] Application of rule "ALL" failed[ERROR:normalizer.cc:213] Failed to verbalize "�13"[ERROR:normalizer.cc:180] Verbalization FAILED for ["13(rthirteen][DEBUG:normalizer.cc:195] Verbalize output: Words
Hello world
```
I think this might be due to fact that parsing of tokenize_classify output is somehow incompatible. Is it the case? is there any tutorial?
**Describe the solution you'd like**
Documentation on how to export NeMo grammars to sparrowhawk as paper suggests.
**Describe alternatives you've considered**
Any pointers to debug the issue, maybe some simple grammar that actually works in export.
| {
"login": "balacoon",
"id": 64325946,
"node_id": "MDEyOk9yZ2FuaXphdGlvbjY0MzI1OTQ2",
"avatar_url": "https://avatars.githubusercontent.com/u/64325946?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/balacoon",
"html_url": "https://github.com/balacoon",
"followers_url": "https://api.github.com/users/balacoon/followers",
"following_url": "https://api.github.com/users/balacoon/following{/other_user}",
"gists_url": "https://api.github.com/users/balacoon/gists{/gist_id}",
"starred_url": "https://api.github.com/users/balacoon/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/balacoon/subscriptions",
"organizations_url": "https://api.github.com/users/balacoon/orgs",
"repos_url": "https://api.github.com/users/balacoon/repos",
"events_url": "https://api.github.com/users/balacoon/events{/privacy}",
"received_events_url": "https://api.github.com/users/balacoon/received_events",
"type": "Organization",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2306/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/2306/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2304 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2304/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2304/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2304/events | https://github.com/NVIDIA/NeMo/pull/2304 | 910,870,342 | MDExOlB1bGxSZXF1ZXN0NjYxMjI3OTQ5 | 2,304 | Added JSON manifest's support to transcribe_speech.py | {
"login": "vsl9",
"id": 4344862,
"node_id": "MDQ6VXNlcjQzNDQ4NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4344862?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vsl9",
"html_url": "https://github.com/vsl9",
"followers_url": "https://api.github.com/users/vsl9/followers",
"following_url": "https://api.github.com/users/vsl9/following{/other_user}",
"gists_url": "https://api.github.com/users/vsl9/gists{/gist_id}",
"starred_url": "https://api.github.com/users/vsl9/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/vsl9/subscriptions",
"organizations_url": "https://api.github.com/users/vsl9/orgs",
"repos_url": "https://api.github.com/users/vsl9/repos",
"events_url": "https://api.github.com/users/vsl9/events{/privacy}",
"received_events_url": "https://api.github.com/users/vsl9/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging 6ffaa42f691b38c7012ee5ff3c8c6efd538c4546 into c8131785cb8980ddb76c088eb32baf91c3c32495 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-2551329ce0f2e7e1a90f3b651cfa199021015c86)\n\n**new alerts:**\n\n* 1 for Unused import"
] | 2021-06-03T21:34:39 | 2021-09-22T21:02:01 | 2021-06-04T15:39:38 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2304",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2304",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2304.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2304.patch",
"merged_at": "2021-06-04T15:39:38"
} | Extended `transcribe_speech.py`
- added support for JSON dataset manifest as input (to transcribe datasets for analysis in SDE)
- modified output format to JSON manifest (to store filenames with ASR transcripts)
Signed-off-by: Vitaly Lavrukhin <[email protected]> | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2304/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/2304/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2303 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2303/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2303/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2303/events | https://github.com/NVIDIA/NeMo/pull/2303 | 910,808,904 | MDExOlB1bGxSZXF1ZXN0NjYxMTc2MDIz | 2,303 | fix_cluster_small_sample | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 1 alert** when merging 8660f43394b1b601d799acc123041db49f050566 into d23ee3e3d5ed2fd17608b736ebae2b415a258a9e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-5d92866e1f3bf62f7a9d6f0212f49c6c64e67338)\n\n**new alerts:**\n\n* 1 for Variable defined multiple times"
] | 2021-06-03T19:55:37 | 2021-06-08T18:31:06 | 2021-06-08T18:31:02 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2303",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2303",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2303.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2303.patch",
"merged_at": "2021-06-08T18:31:02"
} | Signed-off-by: nithinraok <[email protected]> | {
"login": "nithinraok",
"id": 19668129,
"node_id": "MDQ6VXNlcjE5NjY4MTI5",
"avatar_url": "https://avatars.githubusercontent.com/u/19668129?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/nithinraok",
"html_url": "https://github.com/nithinraok",
"followers_url": "https://api.github.com/users/nithinraok/followers",
"following_url": "https://api.github.com/users/nithinraok/following{/other_user}",
"gists_url": "https://api.github.com/users/nithinraok/gists{/gist_id}",
"starred_url": "https://api.github.com/users/nithinraok/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/nithinraok/subscriptions",
"organizations_url": "https://api.github.com/users/nithinraok/orgs",
"repos_url": "https://api.github.com/users/nithinraok/repos",
"events_url": "https://api.github.com/users/nithinraok/events{/privacy}",
"received_events_url": "https://api.github.com/users/nithinraok/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2303/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/2303/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2302 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2302/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2302/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2302/events | https://github.com/NVIDIA/NeMo/pull/2302 | 910,554,624 | MDExOlB1bGxSZXF1ZXN0NjYwOTU4MDQ1 | 2,302 | Fix SqueezeExcite forward for fp16 | {
"login": "Oktai15",
"id": 17337773,
"node_id": "MDQ6VXNlcjE3MzM3Nzcz",
"avatar_url": "https://avatars.githubusercontent.com/u/17337773?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Oktai15",
"html_url": "https://github.com/Oktai15",
"followers_url": "https://api.github.com/users/Oktai15/followers",
"following_url": "https://api.github.com/users/Oktai15/following{/other_user}",
"gists_url": "https://api.github.com/users/Oktai15/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Oktai15/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Oktai15/subscriptions",
"organizations_url": "https://api.github.com/users/Oktai15/orgs",
"repos_url": "https://api.github.com/users/Oktai15/repos",
"events_url": "https://api.github.com/users/Oktai15/events{/privacy}",
"received_events_url": "https://api.github.com/users/Oktai15/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-03T14:33:16 | 2021-11-17T08:50:02 | 2021-06-03T21:56:34 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2302",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2302",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2302.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2302.patch",
"merged_at": null
} | We need to have ability of run inference model with SE blocks in fp16
Signed-off-by: Oktai Tatanov <[email protected]> | {
"login": "Oktai15",
"id": 17337773,
"node_id": "MDQ6VXNlcjE3MzM3Nzcz",
"avatar_url": "https://avatars.githubusercontent.com/u/17337773?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Oktai15",
"html_url": "https://github.com/Oktai15",
"followers_url": "https://api.github.com/users/Oktai15/followers",
"following_url": "https://api.github.com/users/Oktai15/following{/other_user}",
"gists_url": "https://api.github.com/users/Oktai15/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Oktai15/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Oktai15/subscriptions",
"organizations_url": "https://api.github.com/users/Oktai15/orgs",
"repos_url": "https://api.github.com/users/Oktai15/repos",
"events_url": "https://api.github.com/users/Oktai15/events{/privacy}",
"received_events_url": "https://api.github.com/users/Oktai15/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2302/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/2302/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2301 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2301/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2301/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2301/events | https://github.com/NVIDIA/NeMo/pull/2301 | 910,201,974 | MDExOlB1bGxSZXF1ZXN0NjYwNjYwMDU1 | 2,301 | fixed paths in tutorials | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-03T07:32:26 | 2021-08-06T08:00:44 | 2021-06-03T08:23:34 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2301",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2301",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2301.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2301.patch",
"merged_at": "2021-06-03T08:23:34"
} | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2301/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/2301/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2300 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2300/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2300/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2300/events | https://github.com/NVIDIA/NeMo/pull/2300 | 910,029,875 | MDExOlB1bGxSZXF1ZXN0NjYwNTE1MjIz | 2,300 | Fastpitch export | {
"login": "borisfom",
"id": 14189615,
"node_id": "MDQ6VXNlcjE0MTg5NjE1",
"avatar_url": "https://avatars.githubusercontent.com/u/14189615?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/borisfom",
"html_url": "https://github.com/borisfom",
"followers_url": "https://api.github.com/users/borisfom/followers",
"following_url": "https://api.github.com/users/borisfom/following{/other_user}",
"gists_url": "https://api.github.com/users/borisfom/gists{/gist_id}",
"starred_url": "https://api.github.com/users/borisfom/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/borisfom/subscriptions",
"organizations_url": "https://api.github.com/users/borisfom/orgs",
"repos_url": "https://api.github.com/users/borisfom/repos",
"events_url": "https://api.github.com/users/borisfom/events{/privacy}",
"received_events_url": "https://api.github.com/users/borisfom/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 3 alerts** and **fixes 1** when merging e99c2fbea67b64a856bd9d59420ad32249fde875 into d23ee3e3d5ed2fd17608b736ebae2b415a258a9e - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-f7d8f6f310df97f17db8897a9e541b9001de6c60)\n\n**new alerts:**\n\n* 3 for Unused import\n\n**fixed alerts:**\n\n* 1 for Unused import"
] | 2021-06-03T02:20:25 | 2021-06-08T21:02:19 | 2021-06-08T21:02:19 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2300",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2300",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2300.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2300.patch",
"merged_at": "2021-06-08T21:02:19"
} | Dependant on https://github.com/NVIDIA/NeMo/pull/2278
| {
"login": "blisc",
"id": 4763352,
"node_id": "MDQ6VXNlcjQ3NjMzNTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/4763352?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/blisc",
"html_url": "https://github.com/blisc",
"followers_url": "https://api.github.com/users/blisc/followers",
"following_url": "https://api.github.com/users/blisc/following{/other_user}",
"gists_url": "https://api.github.com/users/blisc/gists{/gist_id}",
"starred_url": "https://api.github.com/users/blisc/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/blisc/subscriptions",
"organizations_url": "https://api.github.com/users/blisc/orgs",
"repos_url": "https://api.github.com/users/blisc/repos",
"events_url": "https://api.github.com/users/blisc/events{/privacy}",
"received_events_url": "https://api.github.com/users/blisc/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2300/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/2300/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2299 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2299/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2299/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2299/events | https://github.com/NVIDIA/NeMo/issues/2299 | 909,427,538 | MDU6SXNzdWU5MDk0Mjc1Mzg= | 2,299 | BERT Pre-training | {
"login": "kruthikakr",
"id": 12526620,
"node_id": "MDQ6VXNlcjEyNTI2NjIw",
"avatar_url": "https://avatars.githubusercontent.com/u/12526620?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kruthikakr",
"html_url": "https://github.com/kruthikakr",
"followers_url": "https://api.github.com/users/kruthikakr/followers",
"following_url": "https://api.github.com/users/kruthikakr/following{/other_user}",
"gists_url": "https://api.github.com/users/kruthikakr/gists{/gist_id}",
"starred_url": "https://api.github.com/users/kruthikakr/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/kruthikakr/subscriptions",
"organizations_url": "https://api.github.com/users/kruthikakr/orgs",
"repos_url": "https://api.github.com/users/kruthikakr/repos",
"events_url": "https://api.github.com/users/kruthikakr/events{/privacy}",
"received_events_url": "https://api.github.com/users/kruthikakr/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": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [
{
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
}
] | null | [
"Please check your pytorch lightning version - RC1 required 1.1.5, main requires 1.3+",
"I have nemo-toolkit-1.0.0rc2 .. with pytorch lightning version 1.1.5..getting error as \r\n from pytorch_lightning.plugins import DDPPlugin\r\nImportError: cannot import name 'DDPPlugin'\r\n\r\n\r\nAnd with 1.3+ same error as above",
"@kruthikakr I wasn't able to reproduce it. Could you share full stack trace that shows where the DDPPlugin occurs? It looks like an environment setup issue. As an alternative option you could run inside NeMo container. ",
"python bert_pretraining.py\r\nTraceback (most recent call last):\r\n File \"bert_pretraining.py\", line 18, in <module>\r\n from pytorch_lightning.plugins import DDPPlugin\r\nImportError: cannot import name 'DDPPlugin'\r\n\r\nNemo Installation fro Source\r\ngit clone https://github.com/NVIDIA/NeMo\r\ncd NeMo\r\n./reinstall.sh\r\npytorch lightning version 1.1.5.\r\n",
"This looks like an environmental problem. After running ./reinstall.sh on the main branch, you should have PyTorch Lightning >= 1.3.\r\n\r\nMake sure that source code that you're running matches the version that you installed.",
"@kruthikakr did it work? can this issue be closed?",
"I am facing similar kind of issue here.\r\nWhile running the bert pretraining file\r\n\r\n[NeMo I 2021-07-05 05:26:45 bert_pretraining:28] Config:\r\n name: PretrainingBERTFromText\r\n trainer:\r\n gpus: 1\r\n num_nodes: 1\r\n max_epochs: 5\r\n max_steps: null\r\n accumulate_grad_batches: 1\r\n precision: 16\r\n amp_level: O1\r\n accelerator: ddp\r\n gradient_clip_val: 0.0\r\n log_every_n_steps: 1\r\n val_check_interval: 1.0\r\n checkpoint_callback: false\r\n logger: false\r\n model:\r\n nemo_path: null\r\n only_mlm_loss: false\r\n num_tok_classification_layers: 1\r\n num_seq_classification_layers: 2\r\n max_seq_length: 128\r\n mask_prob: 0.15\r\n short_seq_prob: 0.1\r\n language_model:\r\n pretrained_model_name: bert-base-uncased\r\n lm_checkpoint: null\r\n config:\r\n attention_probs_dropout_prob: 0.1\r\n hidden_act: gelu\r\n hidden_dropout_prob: 0.1\r\n hidden_size: 768\r\n initializer_range: 0.02\r\n intermediate_size: 3072\r\n max_position_embeddings: 512\r\n num_attention_heads: 12\r\n num_hidden_layers: 12\r\n type_vocab_size: 2\r\n vocab_size: 30522\r\n config_file: null\r\n tokenizer:\r\n tokenizer_name: ${model.language_model.pretrained_model_name}\r\n vocab_file: null\r\n tokenizer_model: null\r\n special_tokens:\r\n unk_token: '[UNK]'\r\n sep_token: '[SEP]'\r\n pad_token: '[PAD]'\r\n bos_token: '[CLS]'\r\n mask_token: '[MASK]'\r\n eos_token: '[SEP]'\r\n cls_token: '[CLS]'\r\n train_ds:\r\n data_file: /media/nlp_drive/syed/BERT-NeMo/DeepLearningExamples/PyTorch/LanguageModeling/BERT/NeMo/examples/nlp/language_modeling/download_dir/wikitext-2/train.txt\r\n max_seq_length: ${model.max_seq_length}\r\n mask_prob: ${model.mask_prob}\r\n short_seq_prob: ${model.short_seq_prob}\r\n batch_size: 16\r\n shuffle: true\r\n num_samples: -1\r\n num_workers: 2\r\n drop_last: false\r\n pin_memory: false\r\n validation_ds:\r\n data_file: null\r\n max_seq_length: ${model.max_seq_length}\r\n mask_prob: ${model.mask_prob}\r\n short_seq_prob: ${model.short_seq_prob}\r\n batch_size: 16\r\n shuffle: false\r\n num_samples: -1\r\n num_workers: 2\r\n drop_last: false\r\n pin_memory: false\r\n optim:\r\n name: adamw\r\n lr: 3.0e-05\r\n weight_decay: 0.0\r\n sched:\r\n name: CosineAnnealing\r\n warmup_steps: null\r\n warmup_ratio: 0.1\r\n min_lr: 0.0\r\n last_epoch: -1\r\n exp_manager:\r\n exp_dir: null\r\n name: PretrainingBERTFromText\r\n create_tensorboard_logger: true\r\n create_checkpoint_callback: true\r\n \r\nGPU available: True, used: True\r\nTPU available: False, using: 0 TPU cores\r\nUsing native 16bit precision.\r\n[NeMo I 2021-07-05 05:26:45 exp_manager:216] Experiments will be logged at /media/nlp_drive/syed/BERT-NeMo/DeepLearningExamples/PyTorch/LanguageModeling/BERT/NeMo/examples/nlp/language_modeling/nemo_experiments/PretrainingBERTFromText/2021-07-05_05-26-45\r\n[NeMo I 2021-07-05 05:26:45 exp_manager:563] TensorboardLogger has been set up\r\nTraceback (most recent call last):\r\n File \"bert_pretraining.py\", line 31, in main\r\n bert_model = BERTLMModel(cfg.model, trainer=trainer)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/collections/nlp/models/language_modeling/bert_lm_model.py\", line 63, in __init__\r\n super().__init__(cfg=cfg, trainer=trainer)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/core/classes/modelPT.py\", line 130, in __init__\r\n self.setup_multiple_validation_data(val_data_config=None)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/core/classes/modelPT.py\", line 564, in setup_multiple_validation_data\r\n model_utils.resolve_validation_dataloaders(model=self)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/utils/model_utils.py\", line 242, in resolve_validation_dataloaders\r\n model.setup_validation_data(cfg.validation_ds)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/collections/nlp/models/language_modeling/bert_lm_model.py\", line 199, in setup_validation_data\r\n else self._setup_dataloader(val_data_config)\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/collections/nlp/models/language_modeling/bert_lm_model.py\", line 230, in _setup_dataloader\r\n dataset = BertPretrainingDataset(\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/collections/nlp/data/language_modeling/lm_bert_dataset.py\", line 70, in __init__\r\n sentence_indices, sentence_idx_file, data_dir = load_data_indices(\r\n File \"/anaconda/envs/py38_default/lib/python3.8/site-packages/nemo/collections/nlp/data/data_utils/data_preprocessing.py\", line 468, in load_data_indices\r\n data_dir = data_file[: data_file.rfind('/')]\r\nAttributeError: 'NoneType' object has no attribute 'rfind'\r\n",
"Can you suggest some small dataset which is similar to wikicorpus, so that I can run the bert_pretraining.py file",
"@BakingBrains you need to specify validation file too, changes those arguments to mandatory args \r\nhttps://github.com/NVIDIA/NeMo/pull/2449"
] | 2021-06-02T12:04:03 | 2021-07-08T01:04:40 | 2021-07-08T01:04:40 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hi
I am trying to run BERT Pre- training with sample data with reference to https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/zh/latest/nlp/bert_pretraining.html
Steps followed :
1. examples/nlp/language_modeling/get_wkt2.sh download_dir.
2. Changed the config file NeMo/examples/nlp/language_modeling/conf/bert_pretraining_from_text_config.yaml.
3. after running the python bert_pretraining.py
Getting this error
[NeMo W 2021-06-02 17:24:59 optimizers:46] Apex was not found. Using the lamb optimizer will error out.
[NeMo W 2021-06-02 17:25:01 nemo_logging:349] /home/kruthika/NeMo/venv/lib/python3.6/site-packages/omegaconf/basecontainer.py:232: UserWarning: cfg.pretty() is deprecated and will be removed in a future version.
Use OmegaConf.to_yaml(cfg)
category=UserWarning,
[NeMo I 2021-06-02 17:25:01 bert_pretraining:28] Config:
name: PretrainingBERTFromText
trainer:
gpus: 1
num_nodes: 1
max_epochs: 2
max_steps: null
accumulate_grad_batches: 1
precision: 16
amp_level: O1
accelerator: ddp
gradient_clip_val: 0.0
log_every_n_steps: 1
val_check_interval: 1.0
checkpoint_callback: false
logger: false
model:
nemo_path: null
only_mlm_loss: false
num_tok_classification_layers: 1
num_seq_classification_layers: 2
max_seq_length: 128
mask_prob: 0.15
short_seq_prob: 0.1
language_model:
pretrained_model_name: bert-base-uncased
lm_checkpoint: null
config:
attention_probs_dropout_prob: 0.1
hidden_act: gelu
hidden_dropout_prob: 0.1
hidden_size: 768
initializer_range: 0.02
intermediate_size: 3072
max_position_embeddings: 512
num_attention_heads: 12
num_hidden_layers: 12
type_vocab_size: 2
vocab_size: 30522
config_file: null
tokenizer:
tokenizer_name: ${model.language_model.pretrained_model_name}
vocab_file: null
tokenizer_model: null
special_tokens:
unk_token: '[UNK]'
sep_token: '[SEP]'
pad_token: '[PAD]'
bos_token: '[CLS]'
mask_token: '[MASK]'
eos_token: '[SEP]'
cls_token: '[CLS]'
train_ds:
data_file: /home/kruthika/NeMo/download_dir/wikitext-2/train.txt
max_seq_length: ${model.max_seq_length}
mask_prob: ${model.mask_prob}
short_seq_prob: ${model.short_seq_prob}
batch_size: 16
shuffle: true
num_samples: -1
num_workers: 2
drop_last: false
pin_memory: false
validation_ds:
data_file: /home/kruthika/NeMo/download_dir/wikitext-2/valid.txt
max_seq_length: ${model.max_seq_length}
mask_prob: ${model.mask_prob}
short_seq_prob: ${model.short_seq_prob}
batch_size: 16
shuffle: false
num_samples: -1
num_workers: 2
drop_last: false
pin_memory: false
optim:
name: adamw
lr: 3.0e-05
weight_decay: 0.0
sched:
name: CosineAnnealing
warmup_steps: null
warmup_ratio: 0.1
min_lr: 0.0
last_epoch: -1
exp_manager:
exp_dir: null
name: PretrainingBERTFromText
create_tensorboard_logger: true
create_checkpoint_callback: true
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
Using native 16bit precision.
[NeMo I 2021-06-02 17:25:01 exp_manager:210] Experiments will be logged at /home/kruthika/NeMo/examples/nlp/language_modeling/nemo_experiments/PretrainingBERTFromText/2021-06-02_17-25-01
[NeMo I 2021-06-02 17:25:01 exp_manager:550] TensorboardLogger has been set up
Traceback (most recent call last):
File "bert_pretraining.py", line 30, in main
exp_manager(trainer, cfg.get("exp_manager", None))
File "/home/kruthika/NeMo/nemo/utils/exp_manager.py", line 232, in exp_manager
configure_checkpointing(trainer, log_dir, checkpoint_name, cfg.checkpoint_callback_params)
File "/home/kruthika/NeMo/nemo/utils/exp_manager.py", line 679, in configure_checkpointing
checkpoint_callback = NeMoModelCheckpoint(**params)
File "/home/kruthika/NeMo/nemo/utils/exp_manager.py", line 580, in __init__
super().__init__(**kwargs)
TypeError: __init__() got an unexpected keyword argument 'prefix'
Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
| {
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2299/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/2299/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2298 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2298/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2298/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2298/events | https://github.com/NVIDIA/NeMo/pull/2298 | 909,398,416 | MDExOlB1bGxSZXF1ZXN0NjU5OTc4OTQ5 | 2,298 | update quantization | {
"login": "Slyne",
"id": 6286804,
"node_id": "MDQ6VXNlcjYyODY4MDQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/6286804?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Slyne",
"html_url": "https://github.com/Slyne",
"followers_url": "https://api.github.com/users/Slyne/followers",
"following_url": "https://api.github.com/users/Slyne/following{/other_user}",
"gists_url": "https://api.github.com/users/Slyne/gists{/gist_id}",
"starred_url": "https://api.github.com/users/Slyne/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/Slyne/subscriptions",
"organizations_url": "https://api.github.com/users/Slyne/orgs",
"repos_url": "https://api.github.com/users/Slyne/repos",
"events_url": "https://api.github.com/users/Slyne/events{/privacy}",
"received_events_url": "https://api.github.com/users/Slyne/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-02T11:28:03 | 2021-06-02T16:07:59 | 2021-06-02T16:07:59 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2298",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2298",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2298.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2298.patch",
"merged_at": "2021-06-02T16:07:59"
} | Signed-off-by: slyned <[email protected]>
Add CER support
But the results of speech_to_text_quant_infer_trt.py and speech_to_text_quant_infer.py are a little different.
I've also tried to use speech_to_text_quant_infer_trt.py to do inference with NeMo models downloaded from ngc without adding --qat, but the results also are not consistent with speech_to_text_infer.py.
| {
"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/2298/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/2298/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2295 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2295/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2295/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2295/events | https://github.com/NVIDIA/NeMo/pull/2295 | 908,969,507 | MDExOlB1bGxSZXF1ZXN0NjU5NjA0NjI0 | 2,295 | Adding new Models releases on NGC. | {
"login": "VahidooX",
"id": 23551647,
"node_id": "MDQ6VXNlcjIzNTUxNjQ3",
"avatar_url": "https://avatars.githubusercontent.com/u/23551647?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VahidooX",
"html_url": "https://github.com/VahidooX",
"followers_url": "https://api.github.com/users/VahidooX/followers",
"following_url": "https://api.github.com/users/VahidooX/following{/other_user}",
"gists_url": "https://api.github.com/users/VahidooX/gists{/gist_id}",
"starred_url": "https://api.github.com/users/VahidooX/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/VahidooX/subscriptions",
"organizations_url": "https://api.github.com/users/VahidooX/orgs",
"repos_url": "https://api.github.com/users/VahidooX/repos",
"events_url": "https://api.github.com/users/VahidooX/events{/privacy}",
"received_events_url": "https://api.github.com/users/VahidooX/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-02T01:52:57 | 2021-08-06T08:00:37 | 2021-06-02T04:49:50 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2295",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2295",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2295.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2295.patch",
"merged_at": "2021-06-02T04:49:50"
} | Adding the following models:
https://ngc.nvidia.com/catalog/models/nvidia:nemo:asrlm_en_transformer_large_ls
https://ngc.nvidia.com/catalog/models/nvidia:nemo:stt_en_conformer_ctc_small_ls
https://ngc.nvidia.com/catalog/models/nvidia:nemo:stt_en_conformer_ctc_medium_ls
https://ngc.nvidia.com/catalog/models/nvidia:nemo:stt_en_conformer_ctc_large_ls
Added a new section for language models for ASR in the checkpoint section.
| {
"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/2295/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/2295/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2294 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2294/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2294/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2294/events | https://github.com/NVIDIA/NeMo/pull/2294 | 908,837,758 | MDExOlB1bGxSZXF1ZXN0NjU5NDgxNTU0 | 2,294 | Adjust warning messages | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-01T23:42:37 | 2021-06-02T00:09:56 | 2021-06-02T00:02:13 | MEMBER | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2294",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2294",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2294.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2294.patch",
"merged_at": "2021-06-02T00:02:13"
} | Signed-off-by: Oleksii Kuchaiev <[email protected]> | {
"login": "okuchaiev",
"id": 22035961,
"node_id": "MDQ6VXNlcjIyMDM1OTYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22035961?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/okuchaiev",
"html_url": "https://github.com/okuchaiev",
"followers_url": "https://api.github.com/users/okuchaiev/followers",
"following_url": "https://api.github.com/users/okuchaiev/following{/other_user}",
"gists_url": "https://api.github.com/users/okuchaiev/gists{/gist_id}",
"starred_url": "https://api.github.com/users/okuchaiev/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/okuchaiev/subscriptions",
"organizations_url": "https://api.github.com/users/okuchaiev/orgs",
"repos_url": "https://api.github.com/users/okuchaiev/repos",
"events_url": "https://api.github.com/users/okuchaiev/events{/privacy}",
"received_events_url": "https://api.github.com/users/okuchaiev/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2294/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/2294/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2293 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2293/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2293/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2293/events | https://github.com/NVIDIA/NeMo/pull/2293 | 908,833,000 | MDExOlB1bGxSZXF1ZXN0NjU5NDc3MDYz | 2,293 | ASR improvements | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"@okuchaiev fyi - updating the signature for config update to remove deprecated API call for OmegaConf in ModelPT."
] | 2021-06-01T23:37:37 | 2021-06-02T20:02:43 | 2021-06-02T20:02:40 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2293",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2293",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2293.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2293.patch",
"merged_at": "2021-06-02T20:02:40"
} | # Changelog
- Update Citrinet configs.
- Improved error message when numba is not installed.
- Add tarred dataset support for all supported filetypes.
- Ignore extra parameters for layer-normed LSTM.
- Correct deprecated call signature for OmegaConf. | {
"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/2293/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/2293/timeline | null | null | true |
https://api.github.com/repos/NVIDIA/NeMo/issues/2292 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2292/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2292/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2292/events | https://github.com/NVIDIA/NeMo/pull/2292 | 908,832,519 | MDExOlB1bGxSZXF1ZXN0NjU5NDc2NjIy | 2,292 | Time quarter to | {
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [] | 2021-06-01T23:37:08 | 2021-06-02T20:18:01 | 2021-06-02T20:17:58 | CONTRIBUTOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | false | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2292",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2292",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2292.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2292.patch",
"merged_at": "2021-06-02T20:17:58"
} | {
"login": "yzhang123",
"id": 4204271,
"node_id": "MDQ6VXNlcjQyMDQyNzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4204271?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yzhang123",
"html_url": "https://github.com/yzhang123",
"followers_url": "https://api.github.com/users/yzhang123/followers",
"following_url": "https://api.github.com/users/yzhang123/following{/other_user}",
"gists_url": "https://api.github.com/users/yzhang123/gists{/gist_id}",
"starred_url": "https://api.github.com/users/yzhang123/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/yzhang123/subscriptions",
"organizations_url": "https://api.github.com/users/yzhang123/orgs",
"repos_url": "https://api.github.com/users/yzhang123/repos",
"events_url": "https://api.github.com/users/yzhang123/events{/privacy}",
"received_events_url": "https://api.github.com/users/yzhang123/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2292/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/2292/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2291 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2291/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2291/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2291/events | https://github.com/NVIDIA/NeMo/pull/2291 | 908,756,960 | MDExOlB1bGxSZXF1ZXN0NjU5NDA2NDYw | 2,291 | Neural Clustering model | {
"login": "fayejf",
"id": 36722593,
"node_id": "MDQ6VXNlcjM2NzIyNTkz",
"avatar_url": "https://avatars.githubusercontent.com/u/36722593?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fayejf",
"html_url": "https://github.com/fayejf",
"followers_url": "https://api.github.com/users/fayejf/followers",
"following_url": "https://api.github.com/users/fayejf/following{/other_user}",
"gists_url": "https://api.github.com/users/fayejf/gists{/gist_id}",
"starred_url": "https://api.github.com/users/fayejf/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/fayejf/subscriptions",
"organizations_url": "https://api.github.com/users/fayejf/orgs",
"repos_url": "https://api.github.com/users/fayejf/repos",
"events_url": "https://api.github.com/users/fayejf/events{/privacy}",
"received_events_url": "https://api.github.com/users/fayejf/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"This pull request **introduces 26 alerts** when merging 426bfeebea7528ea52a4ab130076546f2a37c7e6 into 4781d1744765fafdcd1238d8d4a9c3f94057aa23 - [view on LGTM.com](https://lgtm.com/projects/g/NVIDIA/NeMo/rev/pr-5739c24c7c6ab00b4184aecc25110c9ddffdeb5a)\n\n**new alerts:**\n\n* 14 for Unused import\n* 8 for Unused local variable\n* 2 for 'import \\*' may pollute namespace\n* 1 for Illegal raise\n* 1 for Variable defined multiple times"
] | 2021-06-01T22:16:03 | 2023-02-22T18:45:17 | 2021-06-01T22:23:22 | COLLABORATOR | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | true | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/pulls/2291",
"html_url": "https://github.com/NVIDIA/NeMo/pull/2291",
"diff_url": "https://github.com/NVIDIA/NeMo/pull/2291.diff",
"patch_url": "https://github.com/NVIDIA/NeMo/pull/2291.patch",
"merged_at": null
} | {
"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/2291/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/2291/timeline | null | null | true |
|
https://api.github.com/repos/NVIDIA/NeMo/issues/2290 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2290/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2290/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2290/events | https://github.com/NVIDIA/NeMo/issues/2290 | 908,151,900 | MDU6SXNzdWU5MDgxNTE5MDA= | 2,290 | [Question] Citrinet and LM for online ASR | {
"login": "ValeryNikiforov",
"id": 52481844,
"node_id": "MDQ6VXNlcjUyNDgxODQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/52481844?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ValeryNikiforov",
"html_url": "https://github.com/ValeryNikiforov",
"followers_url": "https://api.github.com/users/ValeryNikiforov/followers",
"following_url": "https://api.github.com/users/ValeryNikiforov/following{/other_user}",
"gists_url": "https://api.github.com/users/ValeryNikiforov/gists{/gist_id}",
"starred_url": "https://api.github.com/users/ValeryNikiforov/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/ValeryNikiforov/subscriptions",
"organizations_url": "https://api.github.com/users/ValeryNikiforov/orgs",
"repos_url": "https://api.github.com/users/ValeryNikiforov/repos",
"events_url": "https://api.github.com/users/ValeryNikiforov/events{/privacy}",
"received_events_url": "https://api.github.com/users/ValeryNikiforov/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Per feature normalization will not be sufficient to perform normalization over small buffer sizes, that's why a fixed mean and std was used for QuartzNet. You can try it, but I would not expect great results. \r\n\r\nThere are scripts for LM usage with both those types of models - https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/asr/asr_language_modeling.html\r\n\r\nThere will also be addition of neural rescoring in the recent future. ",
"@titu1994 Thank you. I think that I can't use original std and mean constants because of different shape ([1,64] in notebook, Citrinet requires [1,80]). Can you share fixed mean and std constants for Citrinet models (in case you have it)? \r\n\r\nAlso, I met a problem with notebook modification for Citrinet. After changing\r\n`asr_model = nemo_asr.models.EncDecCTCModel.from_pretrained('QuartzNet15x5Base-En')` (works ok with per feature normalization)\r\nto\r\n`asr_model = nemo_asr.models.ctc_bpe_models.EncDecCTCModelBPE.from_pretrained('stt_en_citrinet_256')` (still using per feature)\r\noutput text is always empty (and `logits.shape[0]` is 0). \r\nCan you please give me some advice? What could I have missed?",
"We haven't precomputed these values for the larger datasets that we currently train on.\r\n\r\nI don't know the timeline, but @jbalam-nv is working on streaming asr notebook, which mostly doesn't require this pre calculated normalization tensor. \r\n \r\nLogits.shape[0] corresponds to batch dimension, how is batch dim 0? ",
"@titu1994 Ok, thank you.\r\nI think I found the problem - after loading Citrinet instead of QuartzNet `timestep_duration` is very large (327.68, this is strange..) and `n_timesteps_overlap = int(frame_overlap / timestep_duration) - 2` equals to -2. \r\nAfter that `logits[self.n_timesteps_overlap:-self.n_timesteps_overlap]` is empty when I call _greedy_decoder. \r\n\r\n`n_timesteps_overlap` calculation takes values from Citrinet config. I will try to fix that problem.\r\nEverything works fine when I switch back to QuartzNet.",
"Citrinet performs 8x downsampling, vs QuartzNet which does 2x downsampling. That would be the source of the problem. The notebook was designed to work with 2x and would need modifications to work with 8x. ",
"@ValeryNikiforov did you succeed in running Citrinet over streaming?",
"@aayush6897 \r\nYes, I decode small chunks with new audio data (+ left side context: 2-4 seconds) and use beamsearch_ngram after that. \r\nFor me, text merging is the main problem. I switched to ctcdecode decoder and it's output token's timestamps helped me a lot.",
"Streaming tutorial that should work with Both Conformer-CTC and Citrinet are now in NeMo: https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/Streaming_ASR.ipynb\r\n\r\nThis is also available as a script for offline long form audio decoding: https://github.com/NVIDIA/NeMo/blob/main/examples/asr/speech_to_text_buffered_infer.py\r\n"
] | 2021-06-01T10:35:39 | 2021-09-24T21:14:06 | 2021-06-09T21:26:09 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello,
Thank you for great toolkit, tutorials and models.
I have some questions:
1. I want to use pretrained Citrinet in Online_ASR_Microphone_Demo instead of QuartzNet. I changed normalization to 'per_feature' and initialized EncDecCTCModelBPE. What else do I need to change for the model to work correctly?
2. Do you plan to release a tutorial for online ASR with LM for Citrinet or QuartzNet? | {
"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/2290/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/2290/timeline | null | completed | false |
https://api.github.com/repos/NVIDIA/NeMo/issues/2289 | https://api.github.com/repos/NVIDIA/NeMo | https://api.github.com/repos/NVIDIA/NeMo/issues/2289/labels{/name} | https://api.github.com/repos/NVIDIA/NeMo/issues/2289/comments | https://api.github.com/repos/NVIDIA/NeMo/issues/2289/events | https://github.com/NVIDIA/NeMo/issues/2289 | 908,033,578 | MDU6SXNzdWU5MDgwMzM1Nzg= | 2,289 | Working with Citrinet-1024 | {
"login": "rose768",
"id": 85157007,
"node_id": "MDQ6VXNlcjg1MTU3MDA3",
"avatar_url": "https://avatars.githubusercontent.com/u/85157007?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rose768",
"html_url": "https://github.com/rose768",
"followers_url": "https://api.github.com/users/rose768/followers",
"following_url": "https://api.github.com/users/rose768/following{/other_user}",
"gists_url": "https://api.github.com/users/rose768/gists{/gist_id}",
"starred_url": "https://api.github.com/users/rose768/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/rose768/subscriptions",
"organizations_url": "https://api.github.com/users/rose768/orgs",
"repos_url": "https://api.github.com/users/rose768/repos",
"events_url": "https://api.github.com/users/rose768/events{/privacy}",
"received_events_url": "https://api.github.com/users/rose768/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | [] | closed | false | null | [] | null | [
"Do you wish to train Citrinet 1024 from scratch ? Or simply fine tune?\r\n\r\nDuring training, it is not possible to have samples longer than 16-20 seconds per sample, no matter the model. It is only during inference that you would be able to pass much longer audio samples. \r\n\r\nThe config for 1024 is not shared since it's describes in the paper, the only thing which would need to be changed is the number of filters in the 512 config provided and change dropout to 0.1. I can share one in this thread just for clarity if needed. \r\n\r\nTo answer your questions -\r\n\r\n1) Filters in CNN scales the width of the model - effectively increasing the capacity of the model due to increase in number of parameters. We see significantly improved wer (roughly 4-6%) when moving from 256 filters to 1024. \r\n\r\n2) This tutorial explains how tokenizarion helps ASR models - https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0/tutorials/asr/08_ASR_with_Subword_Tokenization.ipynb\r\n\r\n3) The model config is provided inside the instantiated model - once you initialize the model, you'll be able to do `cfg = model.cfg` to extract the final model config after training. You can print it out, modify it and then create a new model with the modified config if needed. \r\n\r\nThe training and other config info are available in the paper and the configs provided in the examples/ASR/conf/Citrinet path. \r\n\r\n\r\n",
"> Do you wish to train Citrinet 1024 from scratch ? Or simply fine tune?\r\n> \r\n> During training, it is not possible to have samples longer than 16-20 seconds per sample, no matter the model. It is only during inference that you would be able to pass much longer audio samples.\r\n> \r\n> The config for 1024 is not shared since it's describes in the paper, the only thing which would need to be changed is the number of filters in the 512 config provided and change dropout to 0.1. I can share one in this thread just for clarity if needed.\r\n> \r\n> To answer your questions -\r\n> \r\n> 1. Filters in CNN scales the width of the model - effectively increasing the capacity of the model due to increase in number of parameters. We see significantly improved wer (roughly 4-6%) when moving from 256 filters to 1024.\r\n> 2. This tutorial explains how tokenizarion helps ASR models - https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0/tutorials/asr/08_ASR_with_Subword_Tokenization.ipynb\r\n> 3. The model config is provided inside the instantiated model - once you initialize the model, you'll be able to do `cfg = model.cfg` to extract the final model config after training. You can print it out, modify it and then create a new model with the modified config if needed.\r\n> \r\n> The training and other config info are available in the paper and the configs provided in the examples/ASR/conf/Citrinet path.\r\n\r\nThanks a lot.\r\nwhat should I put in tokenizer at config file in below:\r\n`tokenizer:`\r\n` dir: ??? # path to directory which contains either tokenizer.model (bpe) or vocab.txt (for wpe)`\r\n ` type: ??? # Can be either bpe or wpe`\r\nIn other words, I do not know what should be as tokens here.",
"Please review the ASR with Subword tokenizarion tutorial on how to build your own Tokenizer - https://colab.research.google.com/github/NVIDIA/NeMo/blob/v1.0.0/tutorials/asr/08_ASR_with_Subword_Tokenization.ipynb",
"Hello, Thanks for this amazing citrinet model.\r\nI want to finetune Citrinet model for my data.How can i do the speech recognition fine-tuning ? Is there any end to end instruction set or example ? \r\nThanks in advance\r\n@titu1994 ",
"I'm preparing a tutorial for finetuning Char and subword models on other languages, it should be released in the coming week or two. ",
"Both the ASR tutorials also have* segments for finetuning on same language (eng) in them. ",
"> Both the ASR tutorials also hand segments for finetuning on same language (eng) in them.\r\n\r\nSo you mean current tutorials are enough to finetune the model for my data? Thanks a lot for this quick answer.",
"Yes, the Tutorial 1 and tutorial 8 (char and subword respectively) show how to take a pretrained model and finetuned on AN4 (same language - EN). \r\n\r\nMost of the steps remain the same for finetuning on other languages too. "
] | 2021-06-01T08:17:39 | 2021-06-09T21:27:51 | 2021-06-09T21:27:51 | NONE | {
"total": 0,
"completed": 0,
"percent_completed": 0
} | null | null | null | Hello,
Thanks a lot for the powerful ASR toolkit.
I am new to ASR. I recently started working with QuartzNet15*5 from this [link](https://github.com/NVIDIA/NeMo/blob/main/tutorials/asr/01_ASR_with_NeMo.ipynb):
but I have long duration acoustic dataset so I saw Citrinet paper at this [link](https://arxiv.org/abs/2104.01721) and try to use Citrinet for speech recognition. I call Citrinet1024 model insted of quartznet like below:
`citrinet_model = nemo_asr.models.EncDecCTCModelBPE.from_pretrained(model_name="stt_en_citrinet_1024")
`
but I could not find the config file of stt_en_citrinet_1024 from NEMO GitHub. It has just for 348 and 512 due to this [link](https://github.com/NVIDIA/NeMo/tree/main/examples/asr/conf/citrinet).
Would you please explain me:
1) What do the filters do in each block of convolution layer in Citrinet? for instance Citrinet1024. what is the benefit of more filters?
2) How does Citrinet work and why it uses from tokenization?
2) How should I use citrinet1024 (implementation) and how to set parameters in the config file of Citrinet1024?
**Environment overview**
Environment location: Google Colab
Method of NeMo install: !pip install nemo_toolkit[asr]
NeMo version: 1.0.0
Learning Rate: 1e-3
**Environment details**
OS version : "Ubuntu20.04.3 LTS"
PyTorch version : "1.7.1" | {
"login": "titu1994",
"id": 3048602,
"node_id": "MDQ6VXNlcjMwNDg2MDI=",
"avatar_url": "https://avatars.githubusercontent.com/u/3048602?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/titu1994",
"html_url": "https://github.com/titu1994",
"followers_url": "https://api.github.com/users/titu1994/followers",
"following_url": "https://api.github.com/users/titu1994/following{/other_user}",
"gists_url": "https://api.github.com/users/titu1994/gists{/gist_id}",
"starred_url": "https://api.github.com/users/titu1994/starred{/owner}{/repo}",
"subscriptions_url": "https://api.github.com/users/titu1994/subscriptions",
"organizations_url": "https://api.github.com/users/titu1994/orgs",
"repos_url": "https://api.github.com/users/titu1994/repos",
"events_url": "https://api.github.com/users/titu1994/events{/privacy}",
"received_events_url": "https://api.github.com/users/titu1994/received_events",
"type": "User",
"user_view_type": "public",
"site_admin": false
} | {
"url": "https://api.github.com/repos/NVIDIA/NeMo/issues/2289/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/2289/timeline | null | completed | false |
End of preview. Expand
in Dataset Viewer.
Dataset Card for Dataset Name
This dataset contains 10,000 issues and pull requests along with their associated comments of Nvidia Nemo Github repo.
Dataset Details
Dataset Description
- Curated by: [More Information Needed]
- Funded by [optional]: [More Information Needed]
- Shared by [optional]: [More Information Needed]
- Language(s) (NLP): [More Information Needed]
- License: [More Information Needed]
Dataset Sources [optional]
- Repository: [More Information Needed]
- Paper [optional]: [More Information Needed]
- Demo [optional]: [More Information Needed]
Uses
Direct Use
[More Information Needed]
Out-of-Scope Use
[More Information Needed]
Dataset Structure
[More Information Needed]
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data
Data Collection and Processing
[More Information Needed]
Who are the source data producers?
[More Information Needed]
Annotations [optional]
Annotation process
[More Information Needed]
Who are the annotators?
[More Information Needed]
Personal and Sensitive Information
[More Information Needed]
Bias, Risks, and Limitations
[More Information Needed]
Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
Citation [optional]
BibTeX:
[More Information Needed]
APA:
[More Information Needed]
Glossary [optional]
[More Information Needed]
More Information [optional]
[More Information Needed]
Dataset Card Authors [optional]
[More Information Needed]
Dataset Card Contact
[More Information Needed]
- Downloads last month
- 20