Zhiyu c66b2c9cf1 Add support for nvidia modelopt fp8 kv cache (#3223)

2025-02-22 07:04:58 +08:00

..

configs

Model: Support Qwen 72B RM model. (#3772)

2025-02-21 14:38:21 -08:00

constrained

compatible with new outlines (#3435)

2025-02-10 01:51:30 +08:00

distributed

[Fix] Not skip NVML Check on AMD Platform (#3135)

2025-01-25 21:33:51 -08:00

entrypoints

feat: Add SageMaker support (#3740)

2025-02-21 19:31:09 +08:00

layers

Add support for nvidia modelopt fp8 kv cache (#3223)

2025-02-22 07:04:58 +08:00

lora

[Fix] Fix bugs and refactor codes in lora for better scalability. (#3652)

2025-02-20 11:51:57 -08:00

managers

[bug] fixed batch api for DeepSeek V3/R1 (#3754)

2025-02-21 10:28:16 -08:00

mem_cache

Sanity check to prevent performance regression (#3171)

2025-01-27 12:28:17 -08:00

metrics

Improve metrics, logging, and importing orders (#2992)

2025-01-19 18:36:59 -08:00

model_executor

feat: support flashinfer mla with prefix cache (#3643)

2025-02-18 02:06:43 +08:00

model_loader

Add support for nvidia modelopt fp8 kv cache (#3223)

2025-02-22 07:04:58 +08:00

models

Add support for nvidia modelopt fp8 kv cache (#3223)

2025-02-22 07:04:58 +08:00

openai_api

[Fix] OpenAI API adapter tokenizer encoding (#3432)

2025-02-21 09:24:15 -08:00

sampling

Fix repetition penalty (#3139)

2025-01-25 21:48:58 -08:00

speculative

Support NextN (MTP) speculative decoding for DeepSeek-V3/R1 (#3582)

2025-02-15 05:28:34 +08:00

_custom_ops.py

Roll back to use vllm custom allreduce (#3006)

2025-01-20 04:03:15 -08:00

aio_rwlock.py

A better aio rwlock that guarantees the order (#2547)

2024-12-22 15:44:32 -08:00

conversation.py

[Feature] Support minicpmv v2.6 (#2785)

2025-01-18 14:14:19 -08:00

custom_op.py

Use forward_cuda to execute custom op for hip platform (#3305)

2025-02-05 02:58:17 +00:00

function_call_parser.py

Fix: Runtime error for function calling (#3300)

2025-02-06 20:52:01 -08:00

hf_transformers_utils.py

Model: Support Qwen 2.5 vl (#3258)

2025-02-16 00:58:53 -08:00

mm_utils.py

docs: fix module docstrings and copyright headers (#2077)

2024-11-22 22:16:53 +08:00

model_parallel.py

Make torch TP composable with torchao (#2436)

2024-12-11 04:21:42 -08:00

server_args.py

Support NextN (MTP) speculative decoding for DeepSeek-V3/R1 (#3582)

2025-02-15 05:28:34 +08:00

server.py

Misc fixes for eagle (flush_cache, CPU overhead) (#3014)

2025-01-20 20:27:38 -08:00

torch_memory_saver_adapter.py

Add a new api configure_logging to allow dumping the requests (#2875)

2025-01-13 14:24:00 -08:00

utils.py

feat: support flashinfer mla attention for deepseek v3 (#3550)

2025-02-14 08:50:14 +08:00