Go
Home
Pricing
FAQ
Install
Home
Pricing
FAQ
Install
Login
via GitHub
bigscience-workshop/Megatron-DeepSpeed
Pull Requests
Commits
training_flos
LS/alibi
LS/doc
Lucile/add-eval-only-arg
Lucile/delete_unnecessary_brackets
Lucile/useless-parenthesis
add-valid-data
bitfit
bloom-ds-inference-repos2
bloom-inference-meta
bnb-resume-2x
bseval_harness
cc-concurrency
chpt-conversion-fix
ckptavg
cluster_benchmark
consumed_samples_per_valid_dataset
cyclic_valid_dataloaders
debug_with_new_dataset
dependabot/pip/black-24.3.0
ds_ckpt_reshape-with-layer-norm-auto-sync
ds-version-check
fix-sample-ids
fp32-checkpoint-extraction
gpu-direct
hadyelsahar/main
launch-debug
license
log-grad-norm
lumi_eval
lumi_mtf
main
master
megatron-2.4-ds-pipe
mtf_p3
mtf-multival
new-dataset
no-shuffling-option
nozero_reshape
olruwase/ds_ckpt_reshape
olruwase/sync_layer_norms
prefixbseval
preprocess_from_HF_dataset
rm-duplicate-param-count
samson/spm
scratchpad
self_attention_stable_corby
skip-broken-tests
sync4
t0loading
test-conversion
thomas/add_shared_t5
thomas/evaluate_gpt_on_prefix_lm_loss
thomas/evaluate_gpt_speed_if_we_pass_attention_mask
thomas/fix_installation
thomas/fix_layer_norm
thomas/improve_test_to_test_custom_kernel
thomas/mlm_train_script
thomas/opt
thomas/test_different_layer_norm
tp-ln-debug
tr1-13B
tr8-104B
train-no-eval-restart
training_flos_rebase
training_flos
universal_ckpt_info
universal_to_fp32_checkpoint
val_args
addressed jareds comments
shoeybi
committed
5 years ago
197c132e
added fp16 lm cross entropy to bert
shoeybi
committed
5 years ago
78022005
added fp16 cross entropy loss option for gpt2
shoeybi
committed
5 years ago
22e3c7e6
added fp16 cross entropy loss option for gpt2
shoeybi
committed
5 years ago
acfe848e
Some changes noticed late
Neel Kant
committed
5 years ago
c044f59a
Prune changes to only be related to ICT
Neel Kant
committed
5 years ago
32bb4edc
Merge branch 'realm-mlm' of https://gitlab-master.nvidia.com/ADLR/megatron-lm into realm-mlm
Neel Kant
committed
5 years ago
91158c9b
runs with new log loss but plateaus early
Neel Kant
committed
5 years ago
20895f2c
testing
shoeybi
committed
5 years ago
2ede8235
Merge branch 'master' into realm-mlm
Neel Kant
committed
5 years ago
674814a5
Temporarily modify ict_eval_bm25 to build full wiki index
Neel Kant
committed
5 years ago
8b355d93
Misc changes
Neel Kant
committed
5 years ago
51204a4d
Add bm25 evaluation code
Neel Kant
committed
5 years ago
8409e1c1
Correct indexer seq length
Neel Kant
committed
5 years ago
dfb907fe
Merge branch 'realm-mlm' of https://gitlab-master.nvidia.com/ADLR/megatron-lm into realm-mlm
Neel Kant
committed
5 years ago
ab5c4f92
Correct CrossEntropyLoss
Neel Kant
committed
5 years ago
0e8f4331
Corrected realm example building, misc improvements for async concurrency
Neel Kant
committed
5 years ago
2fd4ea6c
Fix token alignment, add mpu checkpointing, misc training code
Neel Kant
committed
5 years ago
8e22824e
Use Faiss GPU index and report retrieval utility
Neel Kant
committed
5 years ago
8573ab35
Restructure Indexer classes
Neel Kant
committed
5 years ago
e59496bf
Fix autoresume by removing code from indexer
Neel Kant
committed
5 years ago
e0a1caba
Change sync variable to gloo backend
Neel Kant
committed
5 years ago
05ea0cca
Async works for total 8 GPU, indexer debug mode
Neel Kant
committed
5 years ago
a670b6c9
Full cycle of communication complete. Also added BasicIndexBuilder
Neel Kant
committed
5 years ago
5684f904
Reorganize indexer. Things run up to saving model checkpoint and repeating
Neel Kant
committed
5 years ago
d4b00be0
Indexer_async works in theory
Neel Kant
committed
5 years ago
e338e311
Add indexer_async
Neel Kant
committed
5 years ago
edaf2aab
Able to run REALM with terrible index sync
Neel Kant
committed
5 years ago
0f0f60aa
Merge branch 'pip_support' into 'master'
shoeybi
committed
5 years ago
5897a790
update
ekmb
committed
5 years ago
faa462fa
Newer
Older