Search results

Small error in https://docs.sglang.io/SpecForge/basic_usage/benchmarking.html

Status: Open.

#356 In sgl-project/SpecForge;

· bingyang-lei opened on Dec 8, 2025

[Question]What factors contribute to Eagle3's reduced acceleration performance on VLM architectures compared to traditional LLMs?

Status: Open.

#352 In sgl-project/SpecForge;

· 330205812 opened on Dec 5, 2025

[Bug] args.build-dataset-num-proc doesn't work

Status: Open.

#349 In sgl-project/SpecForge;

· C3236455482 opened on Dec 3, 2025

[Bug] latest benchmark run llama3.1-8b mat1 RuntimeError: expected mat1 and mat2 to have the same dtype, but got: float != c10::Half

Status: Open.

#348 In sgl-project/SpecForge;

· Sylvan820 opened on Dec 3, 2025

vlm target model don't have set_aux_hidden_states_layers

Status: Open.

#345 In sgl-project/SpecForge;

· C3236455482 opened on Dec 2, 2025

[Question] Why configs/qwen3-30B-A3B-eagle3.json draft_vocab_size=32000

Status: Open.

#341 In sgl-project/SpecForge;

· huang3eng opened on Dec 1, 2025

Throughput degradation on Qwen3-30B-A3B with EAGLE3

Status: Open.

#339 In sgl-project/SpecForge;

· Zzsf11 opened on Dec 1, 2025

draft train

Status: Open.

#338 In sgl-project/SpecForge;

· Abigbigbig opened on Nov 29, 2025

@Abigbigbig This looks like a different issue from this PR. Let's move to a different issue. I can point you the fix

Status: Open.

#336 In sgl-project/SpecForge;

· Abigbigbig opened on Nov 28, 2025

[Bug] Missing bench_model_speedup.py

Status: Open.

#335 In sgl-project/SpecForge;

· lesser-fullness-G opened on Nov 28, 2025

[Bug] Eagle3 training for gpt-oss-120b fails with OOM

Status: Open.

#326 In sgl-project/SpecForge;

· gopalsarda opened on Nov 25, 2025

EAGLE3 on Qwen2.5-VL / Qwen3-VL shows extremely low accept length (accept_len ≈ 1)

bugSomething isn't workingSomething isn't working

low accuracy

Status: Open.

#310 In sgl-project/SpecForge;

· C3236455482 opened on Nov 19, 2025