huggingface/text-generation-inference

Pull Requests Commits

xpu lora support (#3232)

sywangyi committed 202 days ago

Verified f6005d68

[gaudi] Gemma3 sliding window support (#3280)

sywangyi committed 204 days ago

Verified 429dcd9c

Gaudi: add CI (#3160)

baptistecolle committed 210 days ago

Verified 9f38d930

[gaudi] Refine rope memory, do not need to keep sin/cos cache per layer (#3274)

sywangyi committed 211 days ago

Verified 71990741

Neuron backend fix and patch version 3.3.4 (#3273)

dacorvo committed 215 days ago

Verified 238fbd4d

[gaudi] gemma3 text and vlm model intial support. need to add sliding window support later (#3270)

sywangyi committed 216 days ago

Verified 14ee6e78

doc: fix README (#3271)

dacorvo committed 216 days ago

Verified bd1bdebb

[gaudi] Refine logging for Gaudi warmup (#3222)

regisss committed 216 days ago

Verified f13e28c9

chore: prepare release 3.3.3 (#3269)

dacorvo committed 216 days ago

Verified b4d17f18

[Gaudi] use pad_token_id to pad input id (#3268)

sywangyi committed 218 days ago

Verified 0627983c

[Gaudi] Fix the integration-test issues (#3265)

yuanwu2017 committed 221 days ago

Verified 3752143b

[Gaudi] Enable Qwen3_moe model (#3244)

yuanwu2017 committed 221 days ago

Verified ded4cb52

[gaudi] HuggingFaceM4/idefics2-8b issue fix (#3264)

sywangyi committed 221 days ago

Verified a220e57f

[Gaudi] Remove optimum-habana (#3261)

yuanwu2017 committed 222 days ago

Verified e07056ab

[gaudi] Move the _update_cos_sin_cache into get_cos_sin (#3254)

yuanwu2017 committed 222 days ago

Verified 25fdc5f0

[gaudi] Vlm rebase and issue fix in benchmark test (#3263)

sywangyi committed 222 days ago

Verified 613b8dd6

[gaudi] Perf optimization (#3256)

sywangyi committed 223 days ago

Verified 83947767

Bump neuron SDK version (#3260)

dacorvo committed 224 days ago

Verified 79183d16

Remove useless packages (#3253)

yuanwu2017 committed 231 days ago

Verified 1ff9d185

Prepare for 3.3.2 (#3249)

danieldk committed 235 days ago

Verified 249189d9

[gaudi] Fix the Llama-4-Maverick-17B-128E crash issue (#3246)

yuanwu2017 committed 236 days ago

Verified 6b6e30a6

[Gaudi] Fix the OOM issue of Llama-4-Scout-17B-16E-Instruct (#3245)

yuanwu2017 committed 237 days ago

Verified 70217ac3

fp8 compressed tensors w8a8 support for Gaudi backend (#3242)

sywangyi committed 237 days ago

Verified f1404400

Add Qwen3 for Gaudi backend (#3229)

yuanwu2017 committed 243 days ago

Verified 1883a62a

Nix: switch to hf-nix (#3240)

danieldk committed 243 days ago

Verified f58d7cf5

Upgrade to new vllm extension ops for Gaudi backend (fix issue in exponential bucketing) (#3239)

sywangyi committed 243 days ago

Verified f08b44ad

Prepare for 3.3.1 (#3238)

danieldk committed 244 days ago

Verified 674c514d

Move input_ids to hpu and remove disposal of adapter_meta (#3237)

sywangyi committed 244 days ago

Verified 9e7e5469

Switch to punica-sgmv kernel from the Hub (#3236)

danieldk committed 244 days ago

Verified e3252879

Fix the crash in default ATTENTION path for Gaudi backend (#3235)

sywangyi committed 245 days ago

Verified 43b1b07f

Newer Older