| .. |
|
data_efficiency
|
ef869377e9
DeepSpeed Data Efficiency Library (#2585)
|
1 年之前 |
|
mii
|
b8b98c6046
MII blog post (#2418)
|
2 年之前 |
|
1.3B-MoE-128.png
|
38e16c696d
add moe-inference tutorial (#1706)
|
2 年之前 |
|
175b-trend.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
1cycle_lr.png
|
b84a1fa410
Web edits (#147)
|
4 年之前 |
|
1t-trend.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
3d-parallelism.png
|
65c2f974d8
Pipeline parallel training engine. (#392)
|
4 年之前 |
|
3pillars.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
530b-trend.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
DeepSpeed-vs-Megatron.png
|
f2ac7eafd5
ZeRO-2 (#217)
|
4 年之前 |
|
DeepSpeed_dark.svg
|
c2735996c0
[docs] add logo (#1676)
|
2 年之前 |
|
DeepSpeed_dark_transparent.svg
|
3a4cb04243
[docs] switch to transparent dark logo
|
2 年之前 |
|
DeepSpeed_light.svg
|
c2735996c0
[docs] add logo (#1676)
|
2 年之前 |
|
DeepSpeed_light_transparent.svg
|
2662fded2d
add logo and move news (#1709)
|
2 年之前 |
|
accelerate-dark.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
accelerate-light.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
accelerate.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
adam-convergence.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
auto-tp-chart-latency.png
|
d92539509b
Auto TP Tutorial with T5 Example (#2962)
|
1 年之前 |
|
auto-tp-chart-opt-throughput.png
|
d92539509b
Auto TP Tutorial with T5 Example (#2962)
|
1 年之前 |
|
auto-tp-chart-throughput.png
|
d92539509b
Auto TP Tutorial with T5 Example (#2962)
|
1 年之前 |
|
bert-ib.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
bert-large-training-time.png
|
5fb22a055a
Ported BERT pre-training tutorial (#184)
|
4 年之前 |
|
bert-scaling.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
bert-tcp.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
bingbert-mixedbit.png
|
95fe2c42e0
fix inference titles and add MoQ pictures (#1092)
|
3 年之前 |
|
convergence-table.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
deepspeed-logo-uppercase-bold-white-1.15.svg
|
908d616072
Website posts and tutorial improvements (#1799)
|
2 年之前 |
|
deepspeed-logo-uppercase-bold-white.svg
|
908d616072
Website posts and tutorial improvements (#1799)
|
2 年之前 |
|
deepspeed-logo-uppercase-white.svg
|
908d616072
Website posts and tutorial improvements (#1799)
|
2 年之前 |
|
deepspeed-speedup.png
|
f2ac7eafd5
ZeRO-2 (#217)
|
4 年之前 |
|
determined.svg
|
2d8f3f564d
Add Determined to open-source DL frameworks (#2573)
|
1 年之前 |
|
end-to-end-bert-training.PNG
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
gpu-numbers.png
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
|
hf-logo.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
hf-transformers.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
inference-gemm-scheduling.png
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
|
inference-kernel-fusion.png
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
|
inference-latency.png
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
|
inference-throughput.png
|
ed3de0c21b
Quantization + inference release (#1091)
|
3 年之前 |
|
large-model-graph.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
layernorm_animation.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
layernorm_deepspeed.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
layernorm_ds.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
layernorm_pytorch.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
layernorm_torch.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
lightning-dark.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
lightning-dark.svg
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
lightning-light.svg
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
lightning.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
loss_and_lr.png
|
b84a1fa410
Web edits (#147)
|
4 年之前 |
|
lr_schedule.png
|
b84a1fa410
Web edits (#147)
|
4 年之前 |
|
megatron-gpt2-perf-test.png
|
b84a1fa410
Web edits (#147)
|
4 年之前 |
|
model_convergence.png
|
b84a1fa410
Web edits (#147)
|
4 年之前 |
|
moe-lat-tput.png
|
e27a60a879
Add more context for the MoE Inference tutorial (#1707)
|
2 年之前 |
|
moe-nlg.png
|
3ffeaa4999
MoE for NLG announcement (#1628)
|
2 年之前 |
|
mosaicml.svg
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
old-vs-new-azure.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
onebit-adam-overview.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
onebit-convergence.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
oom_dp8_1.5B_log.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
perf-overview.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
pipe-schedule.png
|
65c2f974d8
Pipeline parallel training engine. (#392)
|
4 年之前 |
|
pp-lowbw-gpt2.png
|
a8a8b3d288
Landing page updates (#395)
|
4 年之前 |
|
prmoe.png
|
38e16c696d
add moe-inference tutorial (#1706)
|
2 年之前 |
|
qkv_fusion.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
quantization-8bit.png
|
95fe2c42e0
fix inference titles and add MoQ pictures (#1092)
|
3 年之前 |
|
quantization-mixedbit.png
|
95fe2c42e0
fix inference titles and add MoQ pictures (#1092)
|
3 年之前 |
|
sa_backward_pass.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_bert_base_time_result.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_bert_large_time_result.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_fixed_sparsity_structure.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_forward_pass.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_gpt2_time_result.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_long_document_comprehension_result.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_maximum_sequence_runnable_on_bert.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
sa_variable_sparsity_structure.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
softmax_animation.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
softmax_deepspeed.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
softmax_ds.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
softmax_pytorch.gif
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
softmax_torch.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
squad-ib.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
squad-scaling.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
squad-tcp.png
|
093f09ff27
Update documentation for 1-bit Adam (#388)
|
4 年之前 |
|
tensorboard_monitor.PNG
|
c87f6ee209
DeepSpeed Monitor Module (Master) (#2013)
|
2 年之前 |
|
transformer_kernel_perf.png
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
transformer_kernel_perf_seq128.PNG
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
transformer_kernel_perf_seq512.PNG
|
734d8991c8
Transformer kernel release (#242)
|
4 年之前 |
|
transformer_preln_arch.png
|
2e6d93e0e5
new transformer pre-ln image (#268)
|
4 年之前 |
|
transformers-dark.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
transformers-light.png
|
a2506b545a
[docs] website refresh (#2123)
|
2 年之前 |
|
variable_sparsity_pattern.png
|
e5bbc2e559
Sparse attn + ops/runtime refactor + v0.3.0 (#343)
|
4 年之前 |
|
vl_moe.png
|
1242d8b4eb
VL MoE Blog (#3120)
|
1 年之前 |
|
vmss-setup.png
|
cf587b8211
DS on Azure blog (#2133)
|
2 年之前 |
|
wandb_monitor.PNG
|
c87f6ee209
DeepSpeed Monitor Module (Master) (#2013)
|
2 年之前 |
|
webinar-aug2020.png
|
7ae8f8bc9b
DeepSpeed webinar announcement (#301)
|
4 年之前 |
|
xtc-1.png
|
0f4f2f982c
Adding DeepSpeed Compression Composer (#2105)
|
2 年之前 |
|
xtc-2.png
|
0f4f2f982c
Adding DeepSpeed Compression Composer (#2105)
|
2 年之前 |
|
xtc-3.png
|
0f4f2f982c
Adding DeepSpeed Compression Composer (#2105)
|
2 年之前 |
|
xtc-4.png
|
0f4f2f982c
Adding DeepSpeed Compression Composer (#2105)
|
2 年之前 |
|
zero-full.png
|
f2ac7eafd5
ZeRO-2 (#217)
|
4 年之前 |
|
zero1_dp8_1.5B_log.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero1_dp8_1.5B_smi.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero2_dp32_10B_log.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero2_dp32_10B_smi.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero3-offload-1-v100.png
|
ba33e86e31
Update ZeRO-Offload tutorials (#824)
|
3 年之前 |
|
zero3-offload-16-v100.png
|
ba33e86e31
Update ZeRO-Offload tutorials (#824)
|
3 年之前 |
|
zero3-offload-200B-scalability.png
|
ba33e86e31
Update ZeRO-Offload tutorials (#824)
|
3 年之前 |
|
zero3-offload-512-v100.png
|
ba33e86e31
Update ZeRO-Offload tutorials (#824)
|
3 年之前 |
|
zero3-offload-memory-overview.png
|
ba33e86e31
Update ZeRO-Offload tutorials (#824)
|
3 年之前 |
|
zero_inference_full_offload.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_model_scale.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_models.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_multi_gpu.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_prefetch.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_token_count_batch_size.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_token_count_cpu_throughput.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_inference_token_count_nvme_throughput.png
|
276eec7beb
ZeRO-Inference blog (#2271)
|
2 年之前 |
|
zero_offload_dp1_10B_cpu.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero_offload_dp1_10B_log.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
|
zero_offload_dp1_10B_smi.png
|
2dea61f285
ZeRO tutorials (#384)
|
4 年之前 |
Molly Smith
