女人自慰AV免费观看内涵网,日韩国产剧情在线观看网址,神马电影网特片网,最新一级电影欧美,在线观看亚洲欧美日韩,黄色视频在线播放免费观看,ABO涨奶期羡澄,第一导航fulione,美女主播操b

0
  • 聊天消息
  • 系統(tǒng)消息
  • 評(píng)論與回復(fù)
登錄后你可以
  • 下載海量資料
  • 學(xué)習(xí)在線課程
  • 觀看技術(shù)視頻
  • 寫文章/發(fā)帖/加入社區(qū)
會(huì)員中心
創(chuàng)作中心

完善資料讓更多小伙伴認(rèn)識(shí)你,還能領(lǐng)取20積分哦,立即完善>

3天內(nèi)不再提示

TensorRT-LLM初探(一)運(yùn)行l(wèi)lama

冬至子 ? 來源:oldpan博客 ? 作者:Oldpan ? 2023-11-16 17:39 ? 次閱讀

前文

TensorRT-LLM正式出來有半個(gè)月了,一直沒有時(shí)間玩,周末趁著有時(shí)間跑一下。

之前玩內(nèi)測(cè)版的時(shí)候就需要cuda-12.x,正式出來仍是需要cuda-12.x,主要是因?yàn)閠ensorr-llm中依賴的CUBIN(二進(jìn)制代碼)是基于cuda12.x編譯生成的,想要跑只能更新驅(qū)動(dòng)。

因此,想要快速跑TensorRT-LLM,建議直接將nvidia-driver升級(jí)到535.xxx,利用docker跑即可,省去自己折騰環(huán)境, 至于想要自定義修改源碼,也在docker中搞就可以

理論上替換原始代碼中的該部分就可以使用別的cuda版本了(batch manager只是不開源,和cuda版本應(yīng)該沒關(guān)系,主要是FMA模塊,另外TensorRT-llm依賴的TensorRT有cuda11.x版本,配合inflight_batcher_llm跑的triton-inference-server也和cuda12.x沒有強(qiáng)制依賴關(guān)系):

image.png

tensorrt-llm中預(yù)先編譯好的部分

說完環(huán)境要求,開始配環(huán)境吧!

搭建運(yùn)行環(huán)境以及庫

首先拉取鏡像,宿主機(jī)顯卡驅(qū)動(dòng)需要高于等于535:

docker pull nvcr.io/nvidia/tritonserver:23.10-trtllm-python-py3

這個(gè)鏡像是前幾天剛出的,包含了運(yùn)行TensorRT-LLM的所有環(huán)境(TensorRT、mpi、nvcc、nccl庫等等),省去自己配環(huán)境的煩惱。

拉下來鏡像后,啟動(dòng)鏡像:

docker run -it -d --cap-add=SYS_PTRACE --cap-add=SYS_ADMIN --security-opt seccomp=unconfined --gpus=all --shm-size=16g --privileged --ulimit memlock=-1 --name=develop nvcr.io/nvidia/tritonserver:23.10-trtllm-python-py3 bash

接下來的操作全在這個(gè)容器里。

編譯tensorrt-llm

首先獲取git倉庫,因?yàn)檫@個(gè)鏡像中 只有運(yùn)行需要的lib ,模型還是需要自行編譯的(因?yàn)橐蕾嚨腡ensorRT,用過trt的都知道需要構(gòu)建engine),所以首先編譯tensorrRT-LLM:

# TensorRT-LLM uses git-lfs, which needs to be installed in advance.
apt-get update && apt-get -y install git git-lfs

git clone https://github.com/NVIDIA/TensorRT-LLM.git
cd TensorRT-LLM
git submodule update --init --recursive
git lfs install
git lfs pull

然后進(jìn)入倉庫進(jìn)行編譯:

python3 ./scripts/build_wheel.py --trt_root /usr/local/tensorrt

一般不會(huì)有環(huán)境問題,這個(gè)docekr中已經(jīng)包含了所有需要的包,執(zhí)行build_wheel的時(shí)候會(huì)按照腳本中的步驟pip install一些需要的包,然后運(yùn)行cmake和make編譯文件:

..
adding 'tensorrt_llm/tools/plugin_gen/templates/functional.py.tpl'
adding 'tensorrt_llm/tools/plugin_gen/templates/plugin.cpp.tpl'
adding 'tensorrt_llm/tools/plugin_gen/templates/plugin.h.tpl'
adding 'tensorrt_llm/tools/plugin_gen/templates/plugin_common.cpp'
adding 'tensorrt_llm/tools/plugin_gen/templates/plugin_common.h'
adding 'tensorrt_llm/tools/plugin_gen/templates/tritonPlugins.cpp.tpl'
adding 'tensorrt_llm-0.5.0.dist-info/LICENSE'
adding 'tensorrt_llm-0.5.0.dist-info/METADATA'
adding 'tensorrt_llm-0.5.0.dist-info/WHEEL'
adding 'tensorrt_llm-0.5.0.dist-info/top_level.txt'
adding 'tensorrt_llm-0.5.0.dist-info/zip-safe'
adding 'tensorrt_llm-0.5.0.dist-info/RECORD'
removing build/bdist.linux-x86_64/wheel
Successfully built tensorrt_llm-0.5.0-py3-none-any.whl

然后pip install tensorrt_llm-0.5.0-py3-none-any.whl即可。

運(yùn)行

首先編譯模型,因?yàn)樽罱鼪]有下載新模型,還是拿舊的llama做例子。其實(shí)吧,其他llm也一樣(chatglm、qwen等等),只要trt-llm支持,編譯運(yùn)行方法都一樣的,在hugging face下載好要測(cè)試的模型即可。

這里我執(zhí)行:

python /work/code/TensorRT-LLM/examples/llama/build.py 
                --model_dir /work/models/GPT/LLAMA/llama-7b-hf   # 可以替換為你自己的llm模型
                --dtype float16 
                --remove_input_padding 
                --use_gpt_attention_plugin float16 
                --enable_context_fmha 
                --use_gemm_plugin float16 
                --use_inflight_batching   # 開啟inflight batching
                --output_dir /work/trtModel/llama/1-gpu

然后就是TensorRT的編譯、構(gòu)建engine的過程(因?yàn)槭褂昧藀lugin,編譯挺快的,這里我只用了一張A4000,所以沒有設(shè)置world_size,默認(rèn)為1),這里有很多細(xì)節(jié),后續(xù)會(huì)聊。

編譯好engine后,會(huì)生成/work/trtModel/llama/1-gpu,后續(xù)會(huì)用到。

執(zhí)行以下命令:

cd tensorrtllm_backend
mkdir triton_model_repo

# 拷貝出來模板模型文件夾
cp -r all_models/inflight_batcher_llm/* triton_model_repo/

# 將剛才生成好的`/work/trtModel/llama/1-gpu`移動(dòng)到模板模型文件夾中
cp /work/trtModel/llama/1-gpu/* triton_model_repo/tensorrt_llm/1

image.png

設(shè)置好之后進(jìn)入tensorrtllm_backend執(zhí)行:

python3 scripts/launch_triton_server.py --world_size=1 --model_repo=triton_model_repo

順利的話就會(huì)輸出:

root@6aaab84e59c0:/work/code/tensorrtllm_backend# I1105 14:16:58.286836 2561098 pinned_memory_manager.cc:241] Pinned memory pool is created at '0x7ffb76000000' with size 268435456
I1105 14:16:58.286973 2561098 cuda_memory_manager.cc:107] CUDA memory pool is created on device 0 with size 67108864
I1105 14:16:58.288120 2561098 model_lifecycle.cc:461] loading: tensorrt_llm:1
I1105 14:16:58.288135 2561098 model_lifecycle.cc:461] loading: preprocessing:1
I1105 14:16:58.288142 2561098 model_lifecycle.cc:461] loading: postprocessing:1
[TensorRT-LLM][WARNING] max_tokens_in_paged_kv_cache is not specified, will use default value
[TensorRT-LLM][WARNING] batch_scheduler_policy parameter was not found or is invalid (must be max_utilization or guaranteed_no_evict)
[TensorRT-LLM][WARNING] kv_cache_free_gpu_mem_fraction is not specified, will use default value of 0.85 or max_tokens_in_paged_kv_cache
[TensorRT-LLM][WARNING] max_num_sequences is not specified, will be set to the TRT engine max_batch_size
[TensorRT-LLM][WARNING] enable_trt_overlap is not specified, will be set to true
[TensorRT-LLM][WARNING] [json.exception.type_error.302] type must be number, but is null
[TensorRT-LLM][WARNING] Optional value for parameter max_num_tokens will not be set.
[TensorRT-LLM][INFO] Initializing MPI with thread mode 1
I1105 14:16:58.392915 2561098 python_be.cc:2199] TRITONBACKEND_ModelInstanceInitialize: postprocessing_0_0 (CPU device 0)
I1105 14:16:58.392979 2561098 python_be.cc:2199] TRITONBACKEND_ModelInstanceInitialize: preprocessing_0_0 (CPU device 0)
[TensorRT-LLM][INFO] MPI size: 1, rank: 0
I1105 14:16:58.732165 2561098 model_lifecycle.cc:818] successfully loaded 'postprocessing'
I1105 14:16:59.383255 2561098 model_lifecycle.cc:818] successfully loaded 'preprocessing'
[TensorRT-LLM][INFO] TRTGptModel maxNumSequences: 16
[TensorRT-LLM][INFO] TRTGptModel maxBatchSize: 8
[TensorRT-LLM][INFO] TRTGptModel enableTrtOverlap: 1
[TensorRT-LLM][INFO] Loaded engine size: 12856 MiB
[TensorRT-LLM][INFO] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 13144, GPU 13111 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] Init cuDNN: CPU +2, GPU +10, now: CPU 13146, GPU 13121 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] TensorRT-managed allocation in engine deserialization: CPU +0, GPU +12852, now: CPU 0, GPU 12852 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 13164, GPU 14363 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] Init cuDNN: CPU +0, GPU +8, now: CPU 13164, GPU 14371 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +0, now: CPU 0, GPU 12852 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] Init cuBLAS/cuBLASLt: CPU +0, GPU +8, now: CPU 13198, GPU 14391 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] Init cuDNN: CPU +0, GPU +10, now: CPU 13198, GPU 14401 (MiB)
[TensorRT-LLM][INFO] [MemUsageChange] TensorRT-managed allocation in IExecutionContext creation: CPU +0, GPU +0, now: CPU 0, GPU 12852 (MiB)
[TensorRT-LLM][INFO] Using 2878 tokens in paged KV cache.
I1105 14:17:17.299293 2561098 model_lifecycle.cc:818] successfully loaded 'tensorrt_llm'
I1105 14:17:17.303661 2561098 model_lifecycle.cc:461] loading: ensemble:1
I1105 14:17:17.305897 2561098 model_lifecycle.cc:818] successfully loaded 'ensemble'
I1105 14:17:17.306051 2561098 server.cc:592] 
+------------------+------+
| Repository Agent | Path |
+------------------+------+
+------------------+------+

I1105 14:17:17.306401 2561098 server.cc:619] 
+-------------+-----------------------------------------------------------------+------------------------------------------------------------------------------------------------------+
| Backend     | Path                                                            | Config                                                                                               |
+-------------+-----------------------------------------------------------------+------------------------------------------------------------------------------------------------------+
| tensorrtllm | /opt/tritonserver/backends/tensorrtllm/libtriton_tensorrtllm.so | {"cmdline":{"auto-complete-config":"false","backend-directory":"/opt/tritonserver/backends","min-com |
|             |                                                                 | pute-capability":"6.000000","default-max-batch-size":"4"}}                                           |
| python      | /opt/tritonserver/backends/python/libtriton_python.so           | {"cmdline":{"auto-complete-config":"false","backend-directory":"/opt/tritonserver/backends","min-com |
|             |                                                                 | pute-capability":"6.000000","shm-region-prefix-name":"prefix0_","default-max-batch-size":"4"}}       |
+-------------+-----------------------------------------------------------------+------------------------------------------------------------------------------------------------------+

I1105 14:17:17.307053 2561098 server.cc:662] 
+----------------+---------+--------+
| Model          | Version | Status |
+----------------+---------+--------+
| ensemble       | 1       | READY  |
| postprocessing | 1       | READY  |
| preprocessing  | 1       | READY  |
| tensorrt_llm   | 1       | READY  |
+----------------+---------+--------+

I1105 14:17:17.393318 2561098 metrics.cc:817] Collecting metrics for GPU 0: NVIDIA RTX A4000
I1105 14:17:17.393534 2561098 metrics.cc:710] Collecting CPU metrics
I1105 14:17:17.394550 2561098 tritonserver.cc:2458] 
+----------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------+
| Option                           | Value                                                                                                                                              |
+----------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------+
| server_id                        | triton                                                                                                                                             |
| server_version                   | 2.39.0                                                                                                                                             |
| server_extensions                | classification sequence model_repository model_repository(unload_dependents) schedule_policy model_configuration system_shared_memory cuda_shared_ |
|                                  | memory binary_tensor_data parameters statistics trace logging                                                                                      |
| model_repository_path[0]         | /work/triton_models/inflight_batcher_llm                                                                                                           |
| model_control_mode               | MODE_NONE                                                                                                                                          |
| strict_model_config              | 1                                                                                                                                                  |
| rate_limit                       | OFF                                                                                                                                                |
| pinned_memory_pool_byte_size     | 268435456                                                                                                                                          |
| cuda_memory_pool_byte_size{0}    | 67108864                                                                                                                                           |
| min_supported_compute_capability | 6.0                                                                                                                                                |
| strict_readiness                 | 1                                                                                                                                                  |
| exit_timeout                     | 30                                                                                                                                                 |
| cache_enabled                    | 0                                                                                                                                                  |
+----------------------------------+----------------------------------------------------------------------------------------------------------------------------------------------------+

I1105 14:17:17.423479 2561098 grpc_server.cc:2513] Started GRPCInferenceService at 0.0.0.0:8001
I1105 14:17:17.424418 2561098 http_server.cc:4497] Started HTTPService at 0.0.0.0:8000

這時(shí)也就啟動(dòng)了triton-inference-server,后端就是TensorRT-LLM。

可以看到LLAMA-7B-FP16精度版本,占用顯存為:

+---------------------------------------------------------------------------------------+
Sun Nov  5 14:20:46 2023       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.113.01             Driver Version: 535.113.01   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA RTX A4000               Off | 00000000:01:00.0 Off |                  Off |
| 41%   34C    P8              16W / 140W |  15855MiB / 16376MiB |      0%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
+---------------------------------------------------------------------------------------+

客戶端

然后我們請(qǐng)求一下吧,先走h(yuǎn)ttp接口

# 執(zhí)行
curl -X POST localhost:8000/v2/models/ensemble/generate -d '{"text_input": "What is machine learning?", "max_tokens": 20, "bad_words": "", "stop_words": ""}'

# 得到返回結(jié)果
{"model_name":"ensemble","model_version":"1","sequence_end":false,"sequence_id":0,"sequence_start":false,"text_output":" ?  What is machine learning? Machine learning is a subfield of computer science that focuses on the development of algorithms that can learn"}

triton目前不支持SSE方法,想stream可以使用grpc協(xié)議,官方也提供了grpc的方法,首先安裝triton客戶端:

pip install tritonclient[all]

然后執(zhí)行:

python3 inflight_batcher_llm/client/inflight_batcher_llm_client.py --request-output-len 200 --tokenizer_dir /work/models/GPT/LLAMA/llama-7b-hf --tokenizer_type llama --streaming

請(qǐng)求后可以看到是一個(gè)token一個(gè)token返回的,也就是我們使用chatgpt3.5時(shí),一個(gè)字一個(gè)字蹦的意思:

... 
[29953]
[29941]
[511]
[450]
[315]
[4664]
[457]
[310]
output_ids =  [[0, 19298, 297, 6641, 29899, 23027, 3444, 29892, 1105, 7598, 16370, 408, 263, 14547, 297, 3681, 1434, 8401, 304, 4517, 297, 29871, 29896, 29947, 29946, 29955, 29889, 940, 3796, 472, 278, 23933, 5977, 322, 278, 7021, 16923, 297, 29258, 265, 1434, 8718, 670, 1914, 27144, 297, 29871, 29896, 29947, 29945, 29896, 29889, 940, 471, 263, 29323, 261, 310, 278, 671, 310, 21837, 7984, 292, 322, 471, 278, 937, 304, 671, 263, 10489, 380, 994, 29889, 940, 471, 884, 263, 410, 29880, 928, 9227, 322, 670, 8277, 5134, 450, 315, 4664, 457, 310, 3444, 313, 29896, 29947, 29945, 29896, 511, 450, 315, 4664, 457, 310, 12730, 313, 29896, 29947, 29945, 29946, 511, 450, 315, 4664, 457, 310, 13616, 313, 29896, 29947, 29945, 29945, 511, 450, 315, 4664, 457, 310, 9556, 313, 29896, 29947, 29945, 29955, 511, 450, 315, 4664, 457, 310, 17362, 313, 29896, 29947, 29945, 29947, 511, 450, 315, 4664, 457, 310, 12710, 313, 29896, 29947, 29945, 29929, 511, 450, 315, 4664, 457, 310, 14198, 653, 313, 29896, 29947, 29953, 29900, 511, 450, 315, 4664, 457, 310, 28806, 313, 29896, 29947, 29953, 29896, 511, 450, 315, 4664, 457, 310, 27440, 313, 29896, 29947, 29953, 29906, 511, 450, 315, 4664, 457, 310, 24506, 313, 29896, 29947, 29953, 29941, 511, 450, 315, 4664, 457, 310]]
Input: Born in north-east France, Soyer trained as a
Output:  chef in Paris before moving to London in 1 847. He worked at the Reform Club and the Royal Hotel in Brighton before opening his own restaurant in 1 851 . He was a pioneer of the use of steam cooking and was the first to use a gas stove. He was also a prolific writer and his books included The Cuisine of France (1 851 ), The Cuisine of Italy (1 854), The Cuisine of Spain (1 855), The Cuisine of Germany (1 857), The Cuisine of Austria (1 858), The Cuisine of Russia (1 859), The Cuisine of Hungary (1 860), The Cuisine of Switzerland (1 861 ), The Cuisine of Norway (1 862), The Cuisine of Sweden (1863), The Cuisine of

因?yàn)殚_了inflight batching,其實(shí)可以同時(shí)多個(gè)請(qǐng)求打過來,修改request_id不要一樣就可以:

# user 1
python3 inflight_batcher_llm/client/inflight_batcher_llm_client.py --request-output-len 200 --tokenizer_dir /work/models/GPT/LLAMA/llama-7b-hf --tokenizer_type llama --streaming --request_id 1
# user 2
python3 inflight_batcher_llm/client/inflight_batcher_llm_client.py --request-output-len 200 --tokenizer_dir /work/models/GPT/LLAMA/llama-7b-hf --tokenizer_type llama --streaming --request_id 2

至此就快速過完整個(gè)TensorRT-LLM的運(yùn)行流程。

使用建議

非常建議使用docker,人生苦短。

在我們實(shí)際使用中,vllm在batch較大的場(chǎng)景并不慢,利用率也能打滿。TensorRT-LLM和vllm的速度在某些模型上快某些模型上慢,各有優(yōu)劣。

image.png

TensorRT-LLM的特點(diǎn)就是借助TensorRT,TensorRT后續(xù)更新越快,支持特性越牛逼,TensorRT-LLM也就越牛逼。靈活性上,我感覺vllm和TensorRT-LLM不分上下,加上大模型的結(jié)構(gòu)其實(shí)都差不多,甚至TensorRT-LLM都沒有上onnx-parser,在后續(xù)更新模型上,python快速搭建模型效率也都差不了多少。

聲明:本文內(nèi)容及配圖由入駐作者撰寫或者入駐合作網(wǎng)站授權(quán)轉(zhuǎn)載。文章觀點(diǎn)僅代表作者本人,不代表電子發(fā)燒友網(wǎng)立場(chǎng)。文章及其配圖僅供工程師學(xué)習(xí)之用,如有內(nèi)容侵權(quán)或者其他違規(guī)問題,請(qǐng)聯(lián)系本站處理。 舉報(bào)投訴
  • python
    +關(guān)注

    關(guān)注

    56

    文章

    4825

    瀏覽量

    86226
  • GPU芯片
    +關(guān)注

    關(guān)注

    1

    文章

    305

    瀏覽量

    6124
  • HTTP接口
    +關(guān)注

    關(guān)注

    0

    文章

    21

    瀏覽量

    1948
  • ChatGPT
    +關(guān)注

    關(guān)注

    29

    文章

    1588

    瀏覽量

    8810
收藏 人收藏

    評(píng)論

    相關(guān)推薦
    熱點(diǎn)推薦

    【算能RADXA微服務(wù)器試用體驗(yàn)】+ GPT語音與視覺交互:1,LLM部署

    。環(huán)境變量的配置,未來在具體項(xiàng)目中我們會(huì)再次提到。 下面我們正式開始項(xiàng)目。項(xiàng)目從輸入到輸出分別涉及了語音識(shí)別,圖像識(shí)別,LLM,TTS這幾個(gè)與AI相關(guān)的模塊。先從最核心的LLM開始。 由于LLAMA
    發(fā)表于 06-25 15:02

    無法在OVMS上運(yùn)行來自Meta的大型語言模型 (LLM),為什么?

    無法在 OVMS 上運(yùn)行來自 Meta 的大型語言模型 (LLM),例如 LLaMa2。 從 OVMS GitHub* 存儲(chǔ)庫運(yùn)行 llama
    發(fā)表于 03-05 08:07

    現(xiàn)已公開發(fā)布!歡迎使用 NVIDIA TensorRT-LLM 優(yōu)化大語言模型推理

    能。該開源程序庫現(xiàn)已作為 NVIDIA NeMo 框架的部分,在 /NVIDIA/TensorRT-LLM GitHub 資源庫中免費(fèi)提供。 大語言模型徹底改變了人工智能領(lǐng)域,并創(chuàng)造了與數(shù)字世界交互
    的頭像 發(fā)表于 10-27 20:05 ?1415次閱讀
    現(xiàn)已公開發(fā)布!歡迎使用 NVIDIA <b class='flag-5'>TensorRT-LLM</b> 優(yōu)化大語言模型推理

    淺析tensorrt-llm搭建運(yùn)行環(huán)境以及庫

    之前玩內(nèi)測(cè)版的時(shí)候就需要cuda-12.x,正式出來仍是需要cuda-12.x,主要是因?yàn)閠ensorr-llm中依賴的CUBIN(二進(jìn)制代碼)是基于cuda12.x編譯生成的,想要跑只能更新驅(qū)動(dòng)。
    的頭像 發(fā)表于 11-13 14:42 ?2779次閱讀
    淺析<b class='flag-5'>tensorrt-llm</b>搭建<b class='flag-5'>運(yùn)行</b>環(huán)境以及庫

    點(diǎn)亮未來:TensorRT-LLM 更新加速 AI 推理性能,支持在 RTX 驅(qū)動(dòng)的 Windows PC 上運(yùn)行新模型

    微軟 Ignite 2023 技術(shù)大會(huì)發(fā)布的新工具和資源包括 OpenAI?Chat?API 的 TensorRT-LLM 封裝接口、RTX 驅(qū)動(dòng)的性能改進(jìn) DirectML?for?Llama?2
    的頭像 發(fā)表于 11-16 21:15 ?854次閱讀
    點(diǎn)亮未來:<b class='flag-5'>TensorRT-LLM</b> 更新加速 AI 推理性能,支持在 RTX 驅(qū)動(dòng)的 Windows PC 上<b class='flag-5'>運(yùn)行</b>新模型

    LLaMA 2是什么?LLaMA 2背后的研究工作

    Meta 發(fā)布的 LLaMA 2,是新的 sota 開源大型語言模型 (LLM)。LLaMA 2 代表著 LLaMA 的下代版本,并且具有
    的頭像 發(fā)表于 02-21 16:00 ?1523次閱讀

    NVIDIA加速微軟最新的Phi-3 Mini開源語言模型

    NVIDIA 宣布使用 NVIDIA TensorRT-LLM 加速微軟最新的 Phi-3 Mini 開源語言模型。TensorRT-LLM個(gè)開源庫,用于優(yōu)化從 PC 到云端的 NVIDIA GPU 上
    的頭像 發(fā)表于 04-28 10:36 ?899次閱讀

    高通支持Meta Llama 3在驍龍終端上運(yùn)行

    高通與Meta攜手合作,共同推動(dòng)Meta的Llama 3大語言模型(LLM)在驍龍驅(qū)動(dòng)的各類終端設(shè)備上實(shí)現(xiàn)高效運(yùn)行。此次合作致力于優(yōu)化Llama 3在智能手機(jī)、個(gè)人電腦、VR/AR頭顯
    的頭像 發(fā)表于 05-09 10:37 ?618次閱讀

    Meta發(fā)布基于Code LlamaLLM編譯器

    近日,科技巨頭Meta在其X平臺(tái)上正式宣布推出了款革命性的LLM編譯器,這模型家族基于Meta Code Llama構(gòu)建,并融合了先進(jìn)的代碼優(yōu)化和編譯器功能。
    的頭像 發(fā)表于 06-29 17:54 ?1799次閱讀

    魔搭社區(qū)借助NVIDIA TensorRT-LLM提升LLM推理效率

    “魔搭社區(qū)是中國最具影響力的模型開源社區(qū),致力給開發(fā)者提供模型即服務(wù)的體驗(yàn)。魔搭社區(qū)利用NVIDIA TensorRT-LLM,大大提高了大語言模型的推理性能,方便了模型應(yīng)用部署,提高了大模型產(chǎn)業(yè)應(yīng)用效率,更大規(guī)模地釋放大模型的應(yīng)用價(jià)值。”
    的頭像 發(fā)表于 08-23 15:48 ?986次閱讀

    TensorRT-LLM低精度推理優(yōu)化

    本文將分享 TensorRT-LLM 中低精度量化內(nèi)容,并從精度和速度角度對(duì)比 FP8 與 INT8。首先介紹性能,包括速度和精度。其次,介紹量化工具 NVIDIA TensorRT Model
    的頭像 發(fā)表于 11-19 14:29 ?1078次閱讀
    <b class='flag-5'>TensorRT-LLM</b>低精度推理優(yōu)化

    使用NVIDIA TensorRT提升Llama 3.2性能

    Llama 3.2 模型集擴(kuò)展了 Meta Llama 開源模型集的模型陣容,包含視覺語言模型(VLM)、小語言模型(SLM)和支持視覺的更新版 Llama Guard 模型。與 NVIDIA 加速
    的頭像 發(fā)表于 11-20 09:59 ?689次閱讀

    NVIDIA TensorRT-LLM Roadmap現(xiàn)已在GitHub上公開發(fā)布

    感謝眾多用戶及合作伙伴直以來對(duì)NVIDIA TensorRT-LLM的支持。TensorRT-LLM 的 Roadmap 現(xiàn)已在 GitHub 上公開發(fā)布!
    的頭像 發(fā)表于 11-28 10:43 ?624次閱讀
    NVIDIA <b class='flag-5'>TensorRT-LLM</b> Roadmap現(xiàn)已在GitHub上公開發(fā)布

    解鎖NVIDIA TensorRT-LLM的卓越性能

    NVIDIA TensorRT-LLM個(gè)專為優(yōu)化大語言模型 (LLM) 推理而設(shè)計(jì)的庫。它提供了多種先進(jìn)的優(yōu)化技術(shù),包括自定義 Attention Kernel、Inflight
    的頭像 發(fā)表于 12-17 17:47 ?718次閱讀

    在NVIDIA TensorRT-LLM中啟用ReDrafter的些變化

    Recurrent Drafting (簡(jiǎn)稱 ReDrafter) 是蘋果公司為大語言模型 (LLM) 推理開發(fā)并開源的種新型推測(cè)解碼技術(shù),該技術(shù)現(xiàn)在可與 NVIDIA TensorRT-LLM
    的頭像 發(fā)表于 12-25 17:31 ?625次閱讀
    在NVIDIA <b class='flag-5'>TensorRT-LLM</b>中啟用ReDrafter的<b class='flag-5'>一</b>些變化