Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / InternLM/lmdeploy issues and pull requests
#2474 - [Bug] The parameters `n=number_outputs` does not work in the function: `response = client.chat.completions.create`
Issue -
State: open - Opened by leoozy 2 days ago
#2473 - Support user-sepcified data type
Pull Request -
State: open - Opened by lvhan028 4 days ago
Labels: enhancement
#2472 - fix [Bug] TypeError: Got unsupported ScalarType BFloat16
Pull Request -
State: open - Opened by SeitaroShinagawa 4 days ago
#2471 - Ascend NPU support
Issue -
State: open - Opened by zer0py2c 4 days ago
- 2 comments
Labels: awaiting response
#2469 - Add silu mul kernel
Pull Request -
State: open - Opened by grimoire 5 days ago
#2468 - [Bug] 2x4090 with Llama2 70B silently crashes (i.e. without any error message in DEBUG mode) as of v0.6.0a0 and v0.6.0 (but works fine in previous versions)
Issue -
State: open - Opened by josephrocca 6 days ago
- 2 comments
#2467 - [Bug] 使用华为昇腾平台 昇腾 910A 显卡推理时报错:Get regInfo failed, The binary_info_config.json of socVersion [ascend910] does not support opType [ApplyRotaryPosEmb].
Issue -
State: open - Opened by XYZliang 6 days ago
- 1 comment
Labels: awaiting response
#2466 - Refactor lora
Pull Request -
State: open - Opened by grimoire 6 days ago
#2465 - Support minicpm3-4b
Pull Request -
State: open - Opened by AllentDan 7 days ago
- 4 comments
Labels: enhancement
#2463 - [Feature] How to use awq with my own dataset
Issue -
State: open - Opened by wangzhongren-code 7 days ago
#2462 - [Bug] fail to deploy model from modelscope using [lmdeploy serve api_server]
Issue -
State: closed - Opened by kaiwang0112006 7 days ago
- 2 comments
#2461 - Fix initialization of runtime_min_p
Pull Request -
State: closed - Opened by irexyc 7 days ago
Labels: Bug:P1
#2460 - fix MultinomialSampling operator builder
Pull Request -
State: closed - Opened by grimoire 7 days ago
Labels: Bug:P2
#2459 - [Bug] The fisrt response is different with others with same GenerationConfig and a given random_seed.
Issue -
State: closed - Opened by zhulinJulia24 7 days ago
- 1 comment
#2458 - [Feature] support s-lora in turbomind backend
Issue -
State: open - Opened by torinchen 8 days ago
- 2 comments
#2457 - [Bug] output not consistent with different max_prefill_token_num for long context input on pytorch engine
Issue -
State: open - Opened by RunningLeon 8 days ago
#2456 - InternVL2在做AWQ量化的时候,不能支持自定义的校准数据集吗
Issue -
State: open - Opened by tanguozhu 8 days ago
- 3 comments
#2455 - MiniCPM3-4B会支持吗?
Issue -
State: open - Opened by LIUKAI0815 8 days ago
- 1 comment
#2454 - fix tensors on different devices when deploying MiniCPM-V-2_6 with tensor parallelism
Pull Request -
State: closed - Opened by irexyc 8 days ago
Labels: Bug:P1
#2453 - [Bug] TypeError: Got unsupported ScalarType BFloat16
Issue -
State: open - Opened by SeitaroShinagawa 8 days ago
- 2 comments
#2452 - [Bug] lmdeploy部署minicpm-v2_6推理报错
Issue -
State: closed - Opened by dfe2342 8 days ago
- 4 comments
#2451 - [Bug] triton.runtime.autotuner.OutOfResources: out of resource: shared memory, Required: 108672, Hardware limit: 101376. Reducing block sizes or `num_stages` may help.
Issue -
State: open - Opened by EvoNexusX 8 days ago
- 3 comments
#2450 - [Bug] LongCite-glm4-9b awq quantization error
Issue -
State: open - Opened by maxin9966 8 days ago
#2449 - support Qwen2-VL with pytorch backend
Pull Request -
State: open - Opened by irexyc 8 days ago
- 3 comments
Labels: enhancement
#2448 - [Feature] pipe如何输出scores
Issue -
State: open - Opened by KooSung 8 days ago
- 1 comment
Labels: awaiting response
#2447 - add docs about ascend
Pull Request -
State: closed - Opened by yao-fengchen 8 days ago
- 1 comment
#2446 - Fix ascend readme
Pull Request -
State: closed - Opened by jinminxi104 9 days ago
- 1 comment
#2445 - bump version to v0.6.0
Pull Request -
State: closed - Opened by lvhan028 9 days ago
- 1 comment
#2444 - fix llama3 rotary in pytorch engine
Pull Request -
State: closed - Opened by grimoire 9 days ago
Labels: Bug:P1
#2443 - [Bug] How to use w4a16 model in PytorchEngine
Issue -
State: closed - Opened by xzmates 9 days ago
- 2 comments
#2442 - [Bug] CUDA runtime error when running Llama-3.1-70B-Instruct-AWQ-INT4
Issue -
State: open - Opened by rtadewald 9 days ago
- 3 comments
Labels: awaiting response
#2441 - [Docs] 自己训练并使用W8A8量化的InternVL v2模型,应该怎么使用python api推理呢?
Issue -
State: closed - Opened by Loneseven 9 days ago
- 3 comments
#2440 - refactor pytorch engine(ascend)
Pull Request -
State: closed - Opened by yao-fengchen 10 days ago
Labels: enhancement
#2439 - [Bug] lmdeploy does not support the regularized lora target module
Issue -
State: open - Opened by orzgugu 10 days ago
- 1 comment
Labels: awaiting response
#2438 - Support pytorch engine kv int4/int8 quantization
Pull Request -
State: open - Opened by AllentDan 10 days ago
- 1 comment
#2437 - [Bug] 请问一下,我用qwen2的模型,仿照internlm-xcomposer2中的plora方式训练了一个VL模型,这个模型应该怎样用lmdeploy部署起来
Issue -
State: open - Opened by alanayu 10 days ago
- 2 comments
Labels: awaiting response
#2436 - [Feature] 能否支持一下qwenvl2
Issue -
State: open - Opened by Ranking666 11 days ago
- 3 comments
Labels: awaiting response
#2435 - [Bug] main分支EngineGenerationConfig不在初始化中了
Issue -
State: closed - Opened by RandomCoins 11 days ago
- 3 comments
#2434 - automatically set max_batch_size according to the device when it is not specified
Pull Request -
State: closed - Opened by lvhan028 12 days ago
Labels: improvement
#2433 - build nccl in dockerfile for cuda11.8
Pull Request -
State: closed - Opened by RunningLeon 13 days ago
Labels: improvement
#2432 - 是否支持embedding模型部署
Issue -
State: open - Opened by Toblame 14 days ago
- 1 comment
#2431 - [ci] regular update
Pull Request -
State: open - Opened by zhulinJulia24 14 days ago
#2430 - [Bug] cogvlm2支持的问题
Issue -
State: open - Opened by tdf1995 14 days ago
- 1 comment
#2429 - lmdeploy测试4卡/8卡部署api_server用benchmark下代码测试发现qps和吞吐没有很大区别
Issue -
State: closed - Opened by sunzx8 14 days ago
- 2 comments
#2428 - Fix some issues encountered by modelscope and community
Pull Request -
State: closed - Opened by irexyc 15 days ago
Labels: Bug:P1
#2427 - inplace logits process as default
Pull Request -
State: closed - Opened by grimoire 15 days ago
Labels: improvement
#2426 - ignore *.pth when download model from model hub
Pull Request -
State: closed - Opened by lvhan028 15 days ago
- 1 comment
Labels: improvement
#2425 - [Bug] 部署internvl2自己lora微调的模型,显存占用非常高,快速爬升到了60g正常吗
Issue -
State: open - Opened by yywangfei 15 days ago
- 1 comment
#2424 - [Feature] Profiling GeMM kernel in lmdeploy
Issue -
State: open - Opened by DerrickYLJ 15 days ago
- 1 comment
#2423 - [Feature] when --tp 2
Issue -
State: open - Opened by maxin9966 15 days ago
- 6 comments
Labels: awaiting response
#2422 - [Docs] AWQ / GPTQ 部分
Issue -
State: open - Opened by Skyseaee 16 days ago
#2421 - build: update ascend dockerfile
Pull Request -
State: closed - Opened by CyCle1024 16 days ago
Labels: improvement
#2420 - support min_p sampling parameter
Pull Request -
State: closed - Opened by irexyc 16 days ago
- 1 comment
Labels: enhancement
#2419 - update actions/download-artifact to v4 to fix security issue
Pull Request -
State: closed - Opened by lvhan028 16 days ago
#2418 - [Docs] 关于kv cache 量化
Issue -
State: closed - Opened by Root970103 16 days ago
- 5 comments
Labels: awaiting response
#2417 - add Ascend get_started
Pull Request -
State: closed - Opened by jinminxi104 16 days ago
Labels: documentation
#2416 - [Bug] Met a error when deploying an AWQ model on H20.
Issue -
State: closed - Opened by medwang1 17 days ago
- 17 comments
#2415 - [Feature] Is there a plan to support the deployment of Qwen2-VL?
Issue -
State: open - Opened by ldknight 17 days ago
- 2 comments
Labels: awaiting response
#2414 - [Bug] CUDA runtime error: operation not supported /lmdeploy/src/turbomind/utils/allocator.h:197
Issue -
State: open - Opened by JingMo 17 days ago
- 12 comments
#2413 - import dlinfer before imageencoding
Pull Request -
State: closed - Opened by jinminxi104 18 days ago
- 3 comments
Labels: improvement
#2412 - [Bug] When using DeepSeek-VL-7B, there is an error raised, because its config doesn't have 'hidden_size'.
Issue -
State: open - Opened by zytx121 18 days ago
- 1 comment
#2411 - [Feature] Would you consider add qwen2vl?
Issue -
State: closed - Opened by PredyDaddy 18 days ago
- 2 comments
#2410 - fix get_started user guide unaccessible
Pull Request -
State: closed - Opened by lvhan028 18 days ago
Labels: documentation
#2409 - [Bug] Aborted (core dumped)
Issue -
State: open - Opened by suwenzhuo 18 days ago
- 1 comment
#2408 - [Bug] 多卡部署InternVL2-8B报错Aborted (core dumped)
Issue -
State: closed - Opened by gxlover0625 19 days ago
- 1 comment
#2407 - [Bug] internlm模型进行bitsandbytes int8量化
Issue -
State: closed - Opened by EvoNexusX 19 days ago
- 14 comments
#2406 - [Feature] vlm开启openai的serve,如何视频解析?
Issue -
State: closed - Opened by maxin9966 19 days ago
#2405 - [Bug] 部署baichuan2-13b-chat报错,是不支持吗?
Issue -
State: open - Opened by sxk000 20 days ago
#2403 - rename the ascend dockerfile
Pull Request -
State: closed - Opened by lvhan028 21 days ago
#2402 - Torchrun launching multiple api_server
Pull Request -
State: open - Opened by AllentDan 21 days ago
#2401 - [ci] add daily test's coverage report
Pull Request -
State: closed - Opened by zhulinJulia24 21 days ago
#2400 - [Bug] 2卡internvl2-26b推理,卡间通信是pcie会失败,nvlink会成功,这是为啥
Issue -
State: open - Opened by chestnut111 21 days ago
- 12 comments
#2399 - [Bug] [TM][ERROR] CUDA runtime error: misaligned address
Issue -
State: open - Opened by sleepwalker2017 21 days ago
- 6 comments
#2398 - 支持对qwen2-audio-instruct的加速吗
Issue -
State: open - Opened by zhanghanweii 21 days ago
- 4 comments
#2397 - [Feature] Does/Can lmdeploy work with XLA/TPUs
Issue -
State: closed - Opened by radna0 22 days ago
- 1 comment
#2396 - fix: make main process exit properly when tp>1 on ascend backend
Pull Request -
State: closed - Opened by CyCle1024 22 days ago
- 1 comment
Labels: Bug:P1
#2395 - Fix /v1/completions batch order wrong
Pull Request -
State: closed - Opened by AllentDan 22 days ago
Labels: Bug:P1
#2394 - [Feature] 增加对于虚拟内存的缓冲时间
Issue -
State: closed - Opened by NB-Group 22 days ago
- 14 comments
#2393 - [Feature] InternVL2 inference is slower than InternLM-Xcomposer2
Issue -
State: closed - Opened by zhaoning1987 23 days ago
- 2 comments
#2392 - Inquiry
Issue -
State: closed - Opened by xiaoajie738 23 days ago
- 2 comments
#2392 - Inquiry
Issue -
State: open - Opened by xiaoajie738 23 days ago
#2391 - [BUG]session id is not threadsafe
Issue -
State: closed - Opened by tp-nan 23 days ago
- 1 comment
#2390 - [Bug] InternLM 2.5 function calling
Issue -
State: open - Opened by coffeecode24 23 days ago
- 3 comments
#2389 - Model Parallel
Issue -
State: closed - Opened by beichenzbc 23 days ago
- 2 comments
#2388 - fix cache position for pytorch engine
Pull Request -
State: closed - Opened by RunningLeon 23 days ago
Labels: Bug:P2
#2388 - fix cache position for pytorch engine
Pull Request -
State: closed - Opened by RunningLeon 23 days ago
Labels: Bug:P2
#2387 - [Bug] error occurs when reset chat session in gradio with image uploded and get response.
Issue -
State: open - Opened by zhulinJulia24 24 days ago
- 14 comments
#2386 - [Feature] 海光DCU简单测试,希望能支持
Issue -
State: open - Opened by luckfu 24 days ago
- 2 comments
#2385 - [Bug] 無法在windows上部署 Phi-3.5-vision-instruct
Issue -
State: open - Opened by HSIAOKUOWEI 24 days ago
- 4 comments
#2385 - [Bug] 無法在windows上部署 Phi-3.5-vision-instruct
Issue -
State: open - Opened by HSIAOKUOWEI 24 days ago
- 4 comments
#2384 - [Bug] 0.6.0 glm4-9b gptq还是会出现无限吐字的问题
Issue -
State: open - Opened by maxin9966 24 days ago
- 20 comments
#2384 - [Bug] 0.6.0 glm4-9b gptq还是会出现无限吐字的问题
Issue -
State: closed - Opened by maxin9966 24 days ago
- 20 comments
#2383 - [Bug] It seems startcode2-7b is not supported, error is PatchedStarcoder2Attention.forward() got an unexpected keyword argument 'cache_position'
Issue -
State: closed - Opened by zhulinJulia24 24 days ago
- 1 comment
#2383 - [Bug] It seems startcode2-7b is not supported, error is PatchedStarcoder2Attention.forward() got an unexpected keyword argument 'cache_position'
Issue -
State: closed - Opened by zhulinJulia24 24 days ago
- 1 comment
#2382 - [Bug] run out of tokens error when using llama3-llava-next-8b-hf
Issue -
State: closed - Opened by binzhang01 24 days ago
- 5 comments
#2381 - [Bug] CUDA runtime error: out of memory /lmdeploy/src/turbomind/utils/memory_utils.cu:32
Issue -
State: open - Opened by AmazDeng 24 days ago
- 5 comments
Labels: awaiting response, Stale
#2381 - [Bug] CUDA runtime error: out of memory /lmdeploy/src/turbomind/utils/memory_utils.cu:32
Issue -
State: closed - Opened by AmazDeng 24 days ago
- 6 comments
Labels: awaiting response, Stale
#2380 - [Feature] Reuse the text prompt encoding when processing batch jobs
Issue -
State: open - Opened by ZejiaZheng 24 days ago
#2379 - [Bug] model format is "gptq" but group_size is 32. Currently, only 128 is supported
Issue -
State: closed - Opened by maxin9966 24 days ago
- 2 comments