microsoft/DeepSpeed

Pull Requests Commits

Add missing Inference sub-configs (#2518)

mrwyattii committed 3 years ago

Verified fe678544

Fix backward compatibility for InferenceConfig (#2516)

mrwyattii committed 3 years ago

Verified e59f8054

Deepspeed quantization library v0.1 (#2450)

lokoppakmsft committed 3 years ago

Verified 78d4ca1f

Add max_tokens alias to max_out_tokens arg to maintain backwards compatibility (#2508)

lekurile committed 3 years ago

Verified d40a15fc

Update docs to autogenerate pydantic config model docs (#2509)

mrwyattii committed 3 years ago

Verified 43bf035c

DeepSpeed inference config. (#2459) (#2472)

awan-10 committed 3 years ago

Verified b5d18a6a

jeffra committed 3 years ago

a4ceabb6

Fix typos: deepseed -> deepspeed (#2499)

jinyouzhi committed 3 years ago

Verified 06e00f61

Make data contiguous before the inplace reshape-copy_ function (#2489)

lokoppakmsft committed 3 years ago

Verified f2710bbe

Fix nightly CI tests (#2493)

mrwyattii committed 3 years ago

Verified be5ec506

Make bf16_optimizer work for non pipeline (#2470)

tjruwase committed 3 years ago

Verified ee39187d

stage_1_and_2.py: no allreduce needed when mp size is 1 (#2494)

guoyejun committed 3 years ago

Verified 3ca9878d

Stable Diffusion Enhancements (#2491)

cmikeh2 committed 3 years ago

Verified e7e75955

Add `scale_attn_by_inverse_layer_idx` feature (#2486)

hyunwoongko committed 3 years ago

Verified 6f77da1b

[docs] add SD tutorial to deepspeed.ai news

jeffra committed 3 years ago

d2d1b4c3

[docs] add SD tutorial to news

jeffra committed 3 years ago

Verified a63cb07e

Fix CI issues related to cupy install (#2483)

mrwyattii committed 3 years ago

Verified 521d329b

Add correct memory-allocation at DeepSpeed-Attention (#2474)

RezaYazdaniAminabadi committed 3 years ago

Verified 9cfcf743

fix accelerate link (#2481)

kyoto7250 committed 3 years ago

Verified a47c3e03

Added MLFLOW environment variables for logging metrics within trainig… (#2477)

savitamittal1 committed 3 years ago

Verified ffb6d987

Updating autotune json default in docs. (#2476)

jomayeri committed 3 years ago

Verified 4a06ecf6

don't gather partitioned activations for mp size 1 (#2454)

guoyejun committed 3 years ago

Verified f74ee318

Create a new folder structure to isolate model-specific code in DS (#2464)

awan-10 committed 3 years ago

Verified 35458da0

fixing the checkpoint loading at inference-engine (#2429)

RezaYazdaniAminabadi committed 3 years ago

Verified 39bdc141

Cache Allocation and Softmax Fixes (#2433)

cmikeh2 committed 3 years ago

Verified 10e9d04c

Fixes for various CI problems (#2457)

mrwyattii committed 3 years ago

Verified 825f9d48

deepspeed/launcher/launch.py: add option '--enable_each_rank_log logdir' (#2409)

guoyejun committed 3 years ago

Verified 3432c740

Reduction Kernel Utility (#2436)

cmikeh2 committed 3 years ago

Verified be4ffb82

Fixing a mismatch in basic adam test. (#2447)

jomayeri committed 3 years ago

Verified 3b3ba3c2

Use CUDA events for inference model profiling (#2371)

mrwyattii committed 3 years ago

Verified e772f166

Newer Older