Skip to content
New issue

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

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

Already on GitHub? Sign in to your account

Merge main to asr_normalize #7084

Merged
merged 123 commits into from
Jul 20, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
123 commits
Select commit Hold shift + click to select a range
89cbf1d
peft eval directly from ckpt (#6785)
arendu Jun 1, 2023
23f1c42
Add Frame-VAD examples and utils (#6463)
stevehuang52 Jun 1, 2023
cfbe092
[TTS][zh] refine hardcoded lowercase for ASCII letters. (#6781)
XuesongYang Jun 2, 2023
5428a97
Spellchecking ASR customization model (#6179)
bene-ges Jun 2, 2023
a420f90
[TTS] Implement new vocoder dataset (#6670)
rlangman Jun 2, 2023
9827c9b
GPT inference long context (#6687)
ekmb Jun 2, 2023
d5819e9
TDT model pull request (#6536)
hainan-xv Jun 2, 2023
ef74006
Fix get_parameters when using main params optimizer (#6764) (#6787)
github-actions[bot] Jun 2, 2023
19a8d2f
Lddl bert (#6761) (#6790)
github-actions[bot] Jun 2, 2023
a7403c2
Fix check (#6798) (#6800)
github-actions[bot] Jun 2, 2023
d984333
Fix validation with drop_last=False (#6704)
mikolajblaz Jun 2, 2023
8f26d83
SDE unt lvl comparison (#6669)
Jorjeous Jun 3, 2023
5c3ed94
Debug Transformer Engine FP8 support with Megatron-core infrastructur…
github-actions[bot] Jun 4, 2023
ae00414
Lora/PEFT training script CI test (#6664)
arendu Jun 5, 2023
76fc488
change branch to main, small fix (#6803)
bene-ges Jun 5, 2023
aa1986f
add call to p2p overlap (#6779) (#6786)
github-actions[bot] Jun 5, 2023
aa21e8a
fixed decor to show messages only when the wrapped object is called.…
XuesongYang Jun 5, 2023
f9bb1b0
Bug fix for reset_sequence_parallel_args (#6802) (#6805)
github-actions[bot] Jun 5, 2023
3063e32
text_generation_utils memory reduction if no logprob needed (#6773)
yzhang123 Jun 6, 2023
010a0e6
Fixed bug in MaskedSpecAug that overestimates samples. (#6775)
tbartley94 Jun 6, 2023
8c26464
update core version (#6817) (#6819)
github-actions[bot] Jun 6, 2023
acf50f4
lora pp2 (#6818)
arendu Jun 6, 2023
04628ca
Add optional index mapping dir in mmap text datasets (#6683)
gheinrich Jun 6, 2023
bf27079
Add inference kv cache support for transformer TE path (#6627)
yen-shi Jun 6, 2023
336372c
Support large inputs to Conformer and Fast Conformer (#6556)
bmwshop Jun 7, 2023
ebfcef7
sharded_manifests updated docs (#6833)
bmwshop Jun 7, 2023
52e23e0
added fc-xl, xxl and titanet-s models (#6832)
nithinraok Jun 8, 2023
6903d9b
add reference to our paper (#6821)
bene-ges Jun 8, 2023
9cca92b
Upperbound Numpy to < 1.24 (#6829)
titu1994 Jun 8, 2023
b67d410
Multi-lookahead cache-aware streaming models (#6711)
VahidooX Jun 8, 2023
3ed3c4e
added changes to ramp up bs (#6799)
dimapihtar Jun 8, 2023
7538a08
Fix typo in core.rst (#6838)
Dounx Jun 8, 2023
014fa02
add back ptuning pp2 test (#6394)
arendu Jun 8, 2023
29ceec0
t5 lora tuning (#6612)
arendu Jun 8, 2023
72faf55
NFA updates (#6695)
erastorgueva-nv Jun 9, 2023
2db352a
Added rouge monitoring support for T5 (#6737)
jubick1337 Jun 10, 2023
a87702a
GPT extrapolatable position embedding (xpos/sandwich/alibi/kerple) an…
hsiehjackson Jun 12, 2023
b54e7fd
fix (#6842) (#6843)
github-actions[bot] Jun 13, 2023
02c3068
Add Frame-VAD to ASR+VAD pipeline (#6464)
stevehuang52 Jun 13, 2023
1e4845c
Add API docs for NeMo Megatron (#6850) (#6864)
github-actions[bot] Jun 14, 2023
72132a2
Update transcribe_utils.py (#6865)
stevehuang52 Jun 15, 2023
a6c8cce
Import Enum for chatbot component (#6877)
ericharper Jun 17, 2023
6f2035b
[bugfix] avoid the random shuffle of phoneme and tone tokens. (#6855)
XuesongYang Jun 17, 2023
990c764
update mcore version (#6875) (#6876)
github-actions[bot] Jun 18, 2023
3aac795
Add trainer.validate example for GPT (#6794) (#6822)
github-actions[bot] Jun 18, 2023
fc84071
typo fix from #6666 (#6882)
arendu Jun 19, 2023
e418f71
Fix k2 build topo helper (#6887)
artbataev Jun 20, 2023
63d9b2c
Update container for import action (#6883)
ericharper Jun 20, 2023
24837af
removed unnecessary print (#6884)
dimapihtar Jun 21, 2023
328bbbb
Fix destructor for delayed mmap dataset case (#6703)
mikolajblaz Jun 21, 2023
07ea971
removed some tests (#6900)
arendu Jun 21, 2023
0b94ef8
Fix transcribe_utils.py for hybrid models in partial transcribe mode …
stevehuang52 Jun 21, 2023
29015df
hot fix SDE (#6897)
Jorjeous Jun 22, 2023
a8609ab
fix ptuning residuals bug (#6866)
arendu Jun 22, 2023
698a5f7
Add hybrid model support to transcribe_speech_parallel.py (#6906)
stevehuang52 Jun 22, 2023
d870644
Make Gradio library optional (#6904)
yidong72 Jun 22, 2023
4726650
Update Frame-VAD doc (#6902)
stevehuang52 Jun 22, 2023
722e77c
Update container info in README.rst (#6913)
fayejf Jun 23, 2023
74cbbb2
Fix fast-glu activation in change partitions (#6909)
hsiehjackson Jun 23, 2023
c4e677a
Documentation for ONNX export of Megatron Models (#6914)
asfiyab-nvidia Jun 25, 2023
f344fdb
FixTextMemMapDataset index file creation in multi-node setup (#6768)
gheinrich Jun 26, 2023
ef56c97
Move model change out of if-branch (#6908)
erastorgueva-nv Jun 26, 2023
e736c86
Update fvad doc (#6920)
stevehuang52 Jun 26, 2023
8204483
Fix flash-attention (#6901)
hsiehjackson Jun 26, 2023
7e3739b
ptuning oom fix (#6916)
arendu Jun 27, 2023
350b2a2
add rampup bs assertion (#6927)
dimapihtar Jun 27, 2023
92c4a2a
Online Code Switching Dataset for ASR (#6579)
trias702 Jun 27, 2023
e9b0b11
Removed optional optimize_for_inference (#6933)
borisfom Jun 28, 2023
7e20750
Apply garbage collection interval to validation steps (#6870) (#6872)
github-actions[bot] Jun 28, 2023
3b4f37a
Enable `rpe` methods in bert-like models (#6898)
sararb Jun 28, 2023
69747d8
Fix AN4 dataset links (#6926)
artbataev Jun 28, 2023
295e88f
Update core commit for CI (#6939)
aklife97 Jun 28, 2023
29b9b8a
Tensor-parallel communication overlap with userbuffer backend (#6792)
github-actions[bot] Jun 28, 2023
5260c9c
lora inference ci (#6931)
arendu Jun 28, 2023
a27ba52
support value attribution condition (#6934)
yidong72 Jun 29, 2023
47c9d74
Fix confidence ensembles RNNT logprobs selection logic for exclude_bl…
KunalDhawan Jun 29, 2023
7e6e04a
Add ub communicator initialization to validation step (#6814)
github-actions[bot] Jun 29, 2023
89225e4
Add missing save restore connector to eval scripts (#6935)
titu1994 Jun 29, 2023
0dee17b
added cache-aware checkpoints. (#6940)
VahidooX Jun 29, 2023
85e3e15
Merge release r1.19.0 into main (#6948)
ericharper Jun 29, 2023
e5706a0
upgrade base container (#6938)
ericharper Jun 30, 2023
b0e5bf3
Fix requirements for pydantic + inflect (#6956)
titu1994 Jul 1, 2023
0b6e4e6
Update distopt API for coalesced NCCL calls (#6886)
timmoon10 Jul 3, 2023
1744718
Remove `compute_on_step` from metrics (#6979) (#6981)
github-actions[bot] Jul 5, 2023
153c307
Fix require_grad typos (#6930)
kit1980 Jul 7, 2023
19449a9
fix the mpt chatbot (#6957) (#6968)
github-actions[bot] Jul 7, 2023
ff430e4
add support for max_total_length=4096 for 43b (#6763)
Zhilin123 Jul 7, 2023
b22a21f
rnnt_greedy_decoding.py: typos? auto-repressively -> auto-regressivel…
vadimkantorov Jul 7, 2023
07b79d2
Cache handling without input tensors mutation (#6980) (#6996)
github-actions[bot] Jul 7, 2023
94e1efa
Hybrid conformer export (#6983) (#6995)
github-actions[bot] Jul 7, 2023
112c806
Fixing an issue with confidence ensembles (#6987) (#7004)
github-actions[bot] Jul 10, 2023
68b4d1f
[TTS] Add cosine distance option to TTS aligner (#6806)
rlangman Jul 11, 2023
0f79a9f
Minor MPT-7B fixes and creation script update (#6982)
trias702 Jul 11, 2023
0cca300
Change Jenkins timeout (#6997)
ericharper Jul 11, 2023
9e75050
remove hard coded input and output fields (#7008)
arendu Jul 11, 2023
41d8477
RoPE length extrapolation with interpolation (#7005)
MaximumEntropy Jul 11, 2023
e87985d
add async + distopt to sft (#7018)
MaximumEntropy Jul 12, 2023
77c666f
Adding tutorial for confidence ensembles (#6932)
Kipok Jul 12, 2023
40c1ccc
Add support for Numba FP16 RNNT Loss (#6991)
titu1994 Jul 12, 2023
a69f0f4
ptuning inference table bug fix (#7015)
arendu Jul 12, 2023
728403d
[TTS] Add tutorial for TTS data prep scripts (#6922)
rlangman Jul 12, 2023
36adc7e
Fix missing import (#7026)
MaximumEntropy Jul 13, 2023
5e02346
fix install_beamsearch_decoders (#7011)
karpnv Jul 13, 2023
caddb8d
Update SDP docs page with a new documentation link (#7029)
Kipok Jul 13, 2023
7ccc2cf
[TTS] Append pretrained FastPitch & SpectrogamEnhancer pair to availa…
github-actions[bot] Jul 13, 2023
d44127e
Add ASR with TTS Tutorial. Fix enhancer usage. (#6955) (#7023)
github-actions[bot] Jul 13, 2023
f7e33fc
Add end_strings to SamplingParams (#6986)
markelsanz14 Jul 13, 2023
5733975
Fix race condition when executing with multi-node where some ranks do…
findkim Jul 13, 2023
470f178
Added bool types to neural_types export (#7032)
tbartley94 Jul 14, 2023
e859e43
rnnt and char utils (#6971)
karpnv Jul 14, 2023
18f283e
fix tab text gen (#7022) (#7031)
github-actions[bot] Jul 14, 2023
33100e0
ASR Confidence update and tutorial (#6810)
GNroy Jul 15, 2023
2ef544f
install_bs (#7019) (#7028)
github-actions[bot] Jul 17, 2023
8b4b382
fixes for spellmapper (#6994) (#7000)
github-actions[bot] Jul 18, 2023
9051440
added back the retro documents (#7033)
yidong72 Jul 18, 2023
84ae944
Remove pyyaml (#7052) (#7054)
github-actions[bot] Jul 18, 2023
47e782a
st standalone model (#6969)
AlexGrinch Jul 18, 2023
ea9d3fd
remove pos emb from state dict for old models (#7068)
ekmb Jul 19, 2023
b1aa4c2
Fix typo in ASR-TTS tutorial (#7049)
artbataev Jul 19, 2023
1dde267
Fixed tutorial's name (#7047)
vsl9 Jul 19, 2023
6704a79
Fix documentation for Numba (#7065) (#7077)
github-actions[bot] Jul 19, 2023
39aff5c
Update Frame-VAD doc and fix onnx export (#7076)
stevehuang52 Jul 19, 2023
d300a3a
memmap worker arg (#7062)
arendu Jul 20, 2023
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
62 changes: 53 additions & 9 deletions .github/workflows/import-test.yml
Original file line number Diff line number Diff line change
Expand Up @@ -6,25 +6,24 @@ on:
paths:
- "**"

# Check https://hub.docker.com/r/pytorch/pytorch/tags for latest tags
jobs:
ci-import-check:
runs-on: ubuntu-latest

# Check https://hub.docker.com/r/pytorch/pytorch/tags for latest tags
test-asr-imports:
runs-on: ubuntu-latest
container:
image: pytorch/pytorch:1.11.0-cuda11.3-cudnn8-runtime

image: pytorch/pytorch:2.0.1-cuda11.7-cudnn8-runtime
steps:
- uses: actions/checkout@v2

- name: Checkout repo
uses: actions/checkout@v2
- name: Update base dependencies
run: |
apt-get update && apt-get install -y build-essential
apt-get install -y libsndfile1 make

- name: Install nemo dependencies
id: nemo-wheel
run: |
pip install Cython
# install test requirements
pip install -r requirements/requirements_test.txt
# Build nemo as a wheel
Expand All @@ -33,7 +32,6 @@ jobs:
# Preserve wheel location
DIST_FILE=$(find ./dist -name "*.whl" | head -n 1)
echo "::set-output name=DIST_FILE::${DIST_FILE}"

- name: Test ASR Domain Imports
run: |
# Install NeMo Domain
Expand All @@ -43,6 +41,29 @@ jobs:
# Uninstall NeMo
pip uninstall -y nemo_toolkit

test-tts-imports:
runs-on: ubuntu-latest
container:
image: pytorch/pytorch:2.0.1-cuda11.7-cudnn8-runtime
steps:
- name: Checkout repo
uses: actions/checkout@v2
- name: Update base dependencies
run: |
apt-get update && apt-get install -y build-essential
apt-get install -y libsndfile1 make
- name: Install nemo dependencies
id: nemo-wheel
run: |
pip install Cython
# install test requirements
pip install -r requirements/requirements_test.txt
# Build nemo as a wheel
pip install build
python -m build --no-isolation --wheel
# Preserve wheel location
DIST_FILE=$(find ./dist -name "*.whl" | head -n 1)
echo "::set-output name=DIST_FILE::${DIST_FILE}"
- name: Test TTS Domain Imports
run: |
# Install NeMo Domain
Expand All @@ -52,6 +73,29 @@ jobs:
# Uninstall NeMo
pip uninstall -y nemo_toolkit

test-nlp-imports:
runs-on: ubuntu-latest
container:
image: pytorch/pytorch:2.0.1-cuda11.7-cudnn8-runtime
steps:
- name: Checkout repo
uses: actions/checkout@v2
- name: Update base dependencies
run: |
apt-get update && apt-get install -y build-essential
apt-get install -y libsndfile1 make
- name: Install nemo dependencies
id: nemo-wheel
run: |
pip install Cython
# install test requirements
pip install -r requirements/requirements_test.txt
# Build nemo as a wheel
pip install build
python -m build --no-isolation --wheel
# Preserve wheel location
DIST_FILE=$(find ./dist -name "*.whl" | head -n 1)
echo "::set-output name=DIST_FILE::${DIST_FILE}"
- name: Test NLP Domain Imports
run: |
# Install NeMo Domain
Expand Down
13 changes: 9 additions & 4 deletions Dockerfile
Original file line number Diff line number Diff line change
Expand Up @@ -14,7 +14,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.

ARG BASE_IMAGE=nvcr.io/nvidia/pytorch:23.04-py3
ARG BASE_IMAGE=nvcr.io/nvidia/pytorch:23.06-py3

# build an image that includes only the nemo dependencies, ensures that dependencies
# are included first for optimal caching, and useful for building a development
Expand Down Expand Up @@ -45,11 +45,11 @@ RUN apt-get update && \
WORKDIR /workspace/

WORKDIR /tmp/
# TODO: Remove once this Apex commit (2/24/23) is included in PyTorch
# TODO: Remove once this Apex commit (5/12/23) is included in PyTorch
# container
RUN git clone https://github.com/NVIDIA/apex.git && \
cd apex && \
git checkout 57057e2fcf1c084c0fcc818f55c0ff6ea1b24ae2 && \
git checkout 8b7a1ff183741dd8f9b87e7bafd04cfde99cea28 && \
pip3 install -v --disable-pip-version-check --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" --global-option="--fast_layer_norm" --global-option="--distributed_adam" --global-option="--deprecated_fused_adam" ./

# uninstall stuff from base container
Expand All @@ -72,6 +72,11 @@ WORKDIR /tmp/nemo
COPY requirements .
RUN for f in $(ls requirements*.txt); do pip3 install --disable-pip-version-check --no-cache-dir -r $f; done

# install flash attention dependencies
RUN pip install flash-attn
# pinned triton version for flash-attention https://github.com/HazyResearch/flash-attention/blob/main/flash_attn/flash_attn_triton.py#L3
RUN pip install triton==2.0.0.dev20221202

# install k2, skip if installation fails
COPY scripts /tmp/nemo/scripts/
RUN INSTALL_MSG=$(/bin/bash /tmp/nemo/scripts/speech_recognition/k2/setup.sh); INSTALL_CODE=$?; \
Expand All @@ -89,7 +94,7 @@ COPY . .

# start building the final container
FROM nemo-deps as nemo
ARG NEMO_VERSION=1.19.0
ARG NEMO_VERSION=1.20.0

# Check that NEMO_VERSION is set. Build will fail without this. Expose NEMO and base container
# version information as runtime environment variable for introspection purposes
Expand Down
Loading
Loading