-
Notifications
You must be signed in to change notification settings - Fork 4.2k
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#5854 In NVIDIA/Megatron-LM;
[Determinism] [MFSDP] Track and verify determinism with MFSDP
enhancementNew feature or requestNew feature or requestStatus: Open.#5852 In NVIDIA/Megatron-LM;- Status: Open.#5844 In NVIDIA/Megatron-LM;
Flaky functional test: hybrid_nemotron_v3_pico_7b_a1b_tp1_ep8_QAD_dgx_h100_1N8G — total loss not bit-exact
bugSomething isn't workingSomething isn't workingStatus: Open.#5832 In NVIDIA/Megatron-LM;Add regression test coverage for get_grad_norm_fp32 (norm_type paths + empty grads_for_norm)
enhancementNew feature or requestNew feature or requestwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#5827 In NVIDIA/Megatron-LM;[inference] Use Triton MLA latent KV append during CUDA Graph capture
enhancementNew feature or requestNew feature or requestwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#5823 In NVIDIA/Megatron-LM;[Inference] Optimize MLA latent KV-cache append with Triton
enhancementNew feature or requestNew feature or requestwaiting-on-maintainersWaiting on maintainers to respondWaiting on maintainers to respondStatus: Open.#5820 In NVIDIA/Megatron-LM;- Status: Open.#5790 In NVIDIA/Megatron-LM;
- Status: Open.#5788 In NVIDIA/Megatron-LM;
- Status: Open.#5787 In NVIDIA/Megatron-LM;
- Status: Open.#5786 In NVIDIA/Megatron-LM;
- Status: Open.#5785 In NVIDIA/Megatron-LM;