All issues

Search results

Compute_rewards in PPO:rewards[j, start:ends[j]][-1] += reward_clip[j] is wrong

Status: Open.

#996In deepspeedai/DeepSpeedExamples;

· CZH-THU opened on Dec 22, 2025

Step1 failed when I run "bash training_scripts/opt/single_gpu/run_1.3b.sh"

Status: Open.

#995In deepspeedai/DeepSpeedExamples;

· niebowen666 opened on Dec 16, 2025

One example, multiple config files

Status: Open.

#989In deepspeedai/DeepSpeedExamples;

· delock opened on Sep 8, 2025

moe example 404

Status: Open.

#984In deepspeedai/DeepSpeedExamples;

· xumj1 opened on Jul 25, 2025

Possible to include an example of DeepNVMe + state dict

Status: Open.

#979In deepspeedai/DeepSpeedExamples;

· sayakpaul opened on Jun 19, 2025

Apply Zero-3 and LoRA appears empty lora weight [0]

Status: Open.

#969In deepspeedai/DeepSpeedExamples;

· jiangxinke opened on Apr 16, 2025

DeepSpeed-FastGen support ascend npu?

Status: Open.

#960In deepspeedai/DeepSpeedExamples;

· RyanOvO opened on Mar 4, 2025

Why Does vf_loss Take the Maximum Value, Rendering Clamp Meaningless?

Status: Open.

#956In deepspeedai/DeepSpeedExamples;

· Morizhaoyang opened on Feb 7, 2025

Is there any example about DeepSpeed Zero with Ulysses/Ulysses-offload

Status: Open.

#951In deepspeedai/DeepSpeedExamples;

· LSC527 opened on Jan 10, 2025

Domino + PP

Status: Open.

#950In deepspeedai/DeepSpeedExamples;

· XZQshiyu opened on Jan 9, 2025

Assertion srcIndex < srcSelectDimSize failed

Status: Open.

#946In deepspeedai/DeepSpeedExamples;

· boqiny opened on Dec 18, 2024

Question to attention computation

Status: Open.

#944In deepspeedai/DeepSpeedExamples;

· yuzhenmao opened on Dec 15, 2024