Commits · 8f241a9645290e5d843e75fabd652c15cee839b1 · OpenDAS / Megatron-LM

18 Nov, 2021 1 commit
- Revert "Update __init__.py" · 8f241a96
  Satpal Singh Rathore authored Nov 18, 2021
```
This reverts commit 2b699d44
```
  8f241a96
17 Nov, 2021 2 commits
- Update utils.py · d03059a6
  Satpal Singh Rathore authored Nov 17, 2021
  
  d03059a6
- Update __init__.py · 2b699d44
  Satpal Singh Rathore authored Nov 17, 2021
```
Removed redundant if-else.
```
  2b699d44
31 Oct, 2021 1 commit
- Typo corrections in README.md · e64507f0
  Satpal Singh Rathore authored Oct 31, 2021
  
  e64507f0
07 Oct, 2021 1 commit
- Merge branch 'fix-inference' into 'main' · b31e1296
  Jared Casper authored Oct 07, 2021
```
Fix inference after T5 pipeline merge

See merge request ADLR/megatron-lm!332
```
  b31e1296
06 Oct, 2021 3 commits
- Merge branch 'slym/jit_arg_typing' into 'main' · cdc614cf
  Jared Casper authored Oct 06, 2021
```
jit function argument type specification

See merge request ADLR/megatron-lm!334
```
  cdc614cf
- Merge branch 'layer-norm-kerne-include' into 'main' · 99559280
  Jared Casper authored Oct 06, 2021
```
THC/THCDeviceUtils.cuh -> ATen/cuda/DeviceUtils.cuh in fused layer norm

See merge request ADLR/megatron-lm!333
```
  99559280
- jit function argument type specification · 838af7d9
  Sangkug Lym authored Oct 05, 2021
  
  838af7d9
05 Oct, 2021 1 commit
- use newer DeviceUtils header · 75e521a0
  Masaki Kozuki authored Oct 06, 2021
  
  75e521a0
01 Oct, 2021 2 commits
- Merge branch 'fix_temp' into 'main' · 386923b5
  Ryan Prenger authored Oct 01, 2021
```
Fixing bug where temperature was never actually broadcast

See merge request ADLR/megatron-lm!330
```
  386923b5
- Fix inference after T5 pipeline merge · f2c35bb0
  Jared Casper authored Oct 01, 2021
```
Adds some backward compatibility code so old inference code still works.
```
  f2c35bb0
30 Sep, 2021 2 commits
- Merge branch 't5_pipeline_parallelism' into 'main' · 5ac5571b
  Jared Casper authored Sep 30, 2021
```
Pipeline parallelism for T5 model

See merge request ADLR/megatron-lm!288
```
  5ac5571b
- Merge branch 'remove-pkg' into 'main' · 9f5eabcf
  Mohammad Shoeybi authored Sep 30, 2021
```
Remove outdated packaging files.

See merge request ADLR/megatron-lm!331
```
  9f5eabcf
29 Sep, 2021 2 commits
- Merge branch 'main' into t5_pipeline_parallelism · cb00a196
  Jared Casper authored Sep 29, 2021
  
  cb00a196
- Remove outdated packaging files. · 6c00f448
  Jared Casper authored Sep 29, 2021
  
  6c00f448
23 Sep, 2021 4 commits
- Had bugs in the fix that I didn't notice until using the big server · 02c00ce6
  rprenger authored Sep 23, 2021
  
  02c00ce6
- Fixing bug where temperature was never actually broadcast · a33e1b35
  rprenger authored Sep 22, 2021
  
  a33e1b35
- Merge branch 'add_BOS' into 'main' · 5ab64637
  Mohammad Shoeybi authored Sep 22, 2021
```
Add Beginning of Sentence token option and adding semaphore while multi-threading to prevent crashes and hangs due to connection keep-alives

See merge request ADLR/megatron-lm!328
```
  5ab64637
- switching from semaphore to lock · 7b293d9b
  rprenger authored Sep 22, 2021
  
  7b293d9b
21 Sep, 2021 3 commits
- Merge branch 'fix_ddp_in_tasks' into 'main' · 14f2c684
  Jared Casper authored Sep 21, 2021
```
Fixing memory bug caused by DDP during task

See merge request ADLR/megatron-lm!329
```
  14f2c684
- addressing comments · f65a0f88
  rprenger authored Sep 21, 2021
  
  f65a0f88
- Fixing memory bug caused by DDP during task · 9d4fd3d3
  rprenger authored Sep 20, 2021
  
  9d4fd3d3
20 Sep, 2021 5 commits
- Fixing merge conflicts · cb57c380
  rprenger authored Sep 20, 2021
  
  cb57c380
- Merge branch 'inference_context_optimization' into 'main' · 87023abd
  Jared Casper authored Sep 20, 2021
```
Inference context optimization

See merge request ADLR/megatron-lm!321
```
  87023abd
- Inference context optimization · 8b9fe87b
  Mohammad Shoeybi authored Sep 20, 2021
  
  8b9fe87b
- Fixing the URL for the web interface · 7bdeb1e7
  rprenger authored Sep 20, 2021
  
  7bdeb1e7
- Adding the option for beginning of sentence token (and fixing hangs) · 69757f9a
  rprenger authored Sep 20, 2021
  
  69757f9a
17 Sep, 2021 3 commits

Merge branch 'add-temperature-parameter-to-server-api' into 'main' · f47aa770
Jared Casper authored Sep 17, 2021
```
Add temperature to the server API

See merge request ADLR/megatron-lm!325
```
f47aa770

Add temperature to the server API · 527e07c0

Robert Clark authored Sep 10, 2021



A temperature value between 0.0 and 100.0 can now be specified via the API
while running the text generation server. The value passed to
--temperature while running the text generation server is kept as the
default value for all API calls that don't include temperature, even if
different values were manually specified previously.
Signed-Off-By: Robert Clark <roclark@nvidia.com>

527e07c0

Merge branch 'fix_initial_broadcasting' into 'main' · a97d676b
Jared Casper authored Sep 17, 2021
```
Fixes a bug in broadcasting that was causing hanging

See merge request ADLR/megatron-lm!327
```
a97d676b

14 Sep, 2021 5 commits
- Merge branch 'debug_harness' into 'main' · 230633f8
  Ryan Prenger authored Sep 14, 2021
```
Changing API to fix LM Harness Evaluation and make API less confusing/terrible

See merge request ADLR/megatron-lm!326
```
  230633f8
- Fixes a bug in broadcasting that was causing hanging · b46482e8
  rprenger authored Sep 14, 2021
  
  b46482e8
- Addressing comments · 593b47b4
  rprenger authored Sep 14, 2021
  
  593b47b4
- Fixing Merge conflict · 2c500111
  rprenger authored Sep 14, 2021
  
  2c500111
- Changing api to tokens_to_generate, making it so we always generate at least tokens_to_generate · 77979e3b
  rprenger authored Sep 14, 2021
  
  77979e3b
13 Sep, 2021 1 commit
- Changing the interface to the lm eval harness and fixing bugs caused by... · 42e83ee0
  rprenger authored Sep 13, 2021
```
Changing the interface to the lm eval harness and fixing bugs caused by misunderstanding out_seq_length
```
  42e83ee0
10 Sep, 2021 1 commit
- Merge branch 'memory_save' into 'main' · c81bf127
  Jared Casper authored Sep 10, 2021
```
Reducing memory usage during inference

See merge request ADLR/megatron-lm!320
```
  c81bf127
08 Sep, 2021 2 commits
- Merge branch 'remove_generate_samples_gpt' into 'main' · 8fe6f9bf
  Jared Casper authored Sep 08, 2021
```
This file doesn't work withthe new inference code, so it should be remove

See merge request ADLR/megatron-lm!322
```
  8fe6f9bf
- This file doesn't work withthe new inference code, so it should be remove · c6f25b30
  rprenger authored Sep 08, 2021
  
  c6f25b30
07 Sep, 2021 1 commit

Merge branch 'slym/allreduce-overlap' into 'main' · b462d254

Jared Casper authored Sep 07, 2021

use low-priority stream for nccl overlapping all-reduce and gemm

See merge request ADLR/megatron-lm!319

b462d254