-
Notifications
You must be signed in to change notification settings - Fork 98
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#828 In pegainfer-project/pegainfer;
- Status: Open.#824 In pegainfer-project/pegainfer;
- Status: Open.#810 In pegainfer-project/pegainfer;
- Status: Open.#807 In pegainfer-project/pegainfer;
- Status: Open.#806 In pegainfer-project/pegainfer;
- Status: Open.#796 In pegainfer-project/pegainfer;
GLM5.2: enable prefix caching with speculative decoding (persist drafter state)
enhancementNew feature or requestNew feature or requestStatus: Open.#792 In pegainfer-project/pegainfer;- Status: Open.#790 In pegainfer-project/pegainfer;
GLM5.2 P/D: apply backpressure when RDMA fetch staging memory is temporarily exhausted
staleAutomatically marked after inactivityAutomatically marked after inactivityStatus: Open.#782 In pegainfer-project/pegainfer;GLM5.2 P/D: distribute TP4 prefill KV publishing beyond rank 0
staleAutomatically marked after inactivityAutomatically marked after inactivityStatus: Open.#781 In pegainfer-project/pegainfer;qwen3: PerToken decode-graph memory grows linearly with batch bucket and is not budgeted
help wantedExtra attention is neededExtra attention is neededhw:1-gpuNeeds one consumer GPU to verifyNeeds one consumer GPU to verifyqwen3Qwen3-4B model crate (pegainfer-qwen3-4b)Qwen3-4B model crate (pegainfer-qwen3-4b)Status: Open.#780 In pegainfer-project/pegainfer;design(pd): integrate OpenInfer P/D workers with Dynamo
staleAutomatically marked after inactivityAutomatically marked after inactivityStatus: Open.#778 In pegainfer-project/pegainfer;