Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / huggingface/optimum-habana issues and pull requests
#971 - fix fsdp lora torch compile issue
Pull Request -
State: closed - Opened by sywangyi 5 months ago
- 7 comments
Labels: run-test, synapse1.16
#970 - Bring flash attention related to Falcon
Pull Request -
State: closed - Opened by libinta 5 months ago
- 1 comment
#969 - eager optimum[habana] pointing to optimum-habana 1.10.4
Issue -
State: closed - Opened by 12010486 5 months ago
- 3 comments
Labels: bug
#968 - Small CI refacto
Pull Request -
State: closed - Opened by regisss 5 months ago
- 1 comment
#967 - Quantization for FSDPA
Pull Request -
State: closed - Opened by dudilester 5 months ago
- 1 comment
Labels: synapse 1.16_dependency
#966 - Remove --fp8 flag from README
Pull Request -
State: closed - Opened by Yantom1 5 months ago
- 1 comment
Labels: synapse1.16
#965 - fix wrongly all_gather for mixtral finetune
Pull Request -
State: closed - Opened by ccrhx4 5 months ago
- 3 comments
Labels: run-test
#964 - Warn user when using --disk_offload without hqt
Pull Request -
State: open - Opened by Yantom1 5 months ago
- 3 comments
Labels: run-test, synapse1.16
#963 - llama70b one card to infer device map with max memory limitation
Pull Request -
State: closed - Opened by Yantom1 5 months ago
- 2 comments
Labels: run-test, synapse1.16
#962 - Mark scale as const and remove --fp8 flag usage
Pull Request -
State: open - Opened by Yantom1 5 months ago
- 6 comments
Labels: run-test, synapse1.16
#961 - Major indexing bug in eager mode
Issue -
State: open - Opened by mattolson93 5 months ago
- 3 comments
Labels: bug
#960 - Fix warmup for diffusers when batch size < throughput_warmup_steps
Pull Request -
State: open - Opened by dsocek 5 months ago
- 1 comment
Labels: run-test, synapse1.16
#959 - Modify gpt_neox accuracy baseline
Pull Request -
State: closed - Opened by mandy-li 5 months ago
- 3 comments
Labels: run-test
#958 - Dev/dlester/fsdpa foh
Pull Request -
State: closed - Opened by dudilester 5 months ago
#957 - Mark scale as const and remove --fp8 flag usage
Pull Request -
State: closed - Opened by Yantom1 5 months ago
#956 - Habana ran
Pull Request -
State: closed - Opened by rkumar2patel 5 months ago
- 1 comment
#955 - Updates run_lora_clm.py with enhanced dataset support
Pull Request -
State: closed - Opened by dmsuehir 5 months ago
- 8 comments
#954 - Warn user when using --disk_offload without hqt
Pull Request -
State: closed - Opened by Yantom1 5 months ago
#953 - Update speech recognition seq2seq example
Pull Request -
State: closed - Opened by regisss 5 months ago
- 1 comment
#952 - [PEFT on Gaudi2C] speed of Full-parameter Finetuning is almost equal to that of LoRA
Issue -
State: closed - Opened by intelyoungway 5 months ago
- 6 comments
#951 - Enable fp8 inference for Llava-hf 7B and 13B in 1.16 release
Pull Request -
State: closed - Opened by Luca-Calabria 5 months ago
- 10 comments
Labels: run-test, synapse 1.16_dependency, synapse1.16
#950 - Update README logo
Pull Request -
State: closed - Opened by regisss 5 months ago
- 1 comment
#949 - Update test generation config to raise `ValueError`
Pull Request -
State: closed - Opened by malkomes 5 months ago
- 1 comment
Labels: run-test
#948 - Fix decoder only generation
Pull Request -
State: open - Opened by tjs-intel 5 months ago
- 9 comments
#947 - Add leaderboard link
Pull Request -
State: closed - Opened by echarlaix 5 months ago
- 1 comment
#946 - Workarounds for failures in using HF Transformers auto classes in Classification sample
Pull Request -
State: open - Opened by sandeep-maddipatla 5 months ago
- 2 comments
#945 - Fix RoPE data type issue for gpt_neox and stablelm (#177)
Pull Request -
State: open - Opened by mandy-li 5 months ago
- 4 comments
Labels: run-test, synapse 1.16_dependency
#944 - Fix formatting of arg parse help strings in the PEFT example
Pull Request -
State: closed - Opened by dmsuehir 5 months ago
- 1 comment
Labels: run-test
#943 - Add support for contrastive search
Pull Request -
State: closed - Opened by skavulya 5 months ago
- 14 comments
Labels: run-test, synapse1.17
#942 - Prefill kvcache upstream
Pull Request -
State: closed - Opened by puneeshkhanna 5 months ago
- 3 comments
#941 - Add information to the audio-classification examples README about --ddp_find_unused_parameters parameter
Pull Request -
State: closed - Opened by Alberto-Villarreal 5 months ago
- 2 comments
Labels: run-test
#940 - use new Habana llama and falcon model configs
Pull Request -
State: closed - Opened by skaulintel 5 months ago
- 1 comment
Labels: run-test
#939 - Fix ValueError for test_summarization
Pull Request -
State: closed - Opened by Jianhong-Zhang 5 months ago
- 1 comment
Labels: run-test
#938 - Grad norm tensor fix
Pull Request -
State: closed - Opened by yeonsily 5 months ago
- 2 comments
Labels: run-test
#937 - Fix wav2vec test load adapter
Pull Request -
State: open - Opened by malkomes 5 months ago
#936 - Initial enablement with FP8 Training (port from OHF #91)
Pull Request -
State: closed - Opened by libinta 5 months ago
- 1 comment
Labels: run-test, synapse1.16
#935 - Add support for Mistral fp8
Pull Request -
State: closed - Opened by jiminha 5 months ago
#934 - Added Llama-2 fp8 text-generation test cases
Pull Request -
State: closed - Opened by yeonsily 5 months ago
- 2 comments
Labels: run-test
#933 - RuntimeError: a Tensor with 32768 elements cannot be converted to Scalar
Issue -
State: closed - Opened by astachowiczhabana 5 months ago
- 3 comments
Labels: bug
#932 - Support InstantX/InstantID
Pull Request -
State: closed - Opened by alekseyfa 5 months ago
- 5 comments
#931 - Support Mistral 32K input token
Pull Request -
State: closed - Opened by jiminha 5 months ago
- 7 comments
Labels: run-test, synapse1.16
#928 - Add Gaudi CI for Sentence Transformers
Pull Request -
State: closed - Opened by regisss 5 months ago
- 1 comment
#926 - sasarkar/Map list to tensors
Pull Request -
State: closed - Opened by ssarkar2 5 months ago
- 1 comment
Labels: run-test, synapse1.16
#925 - fix performance issue in mistral
Pull Request -
State: closed - Opened by xin3he 5 months ago
- 4 comments
#924 - Skaulintel/ci mistral fp8
Pull Request -
State: closed - Opened by skaulintel 5 months ago
- 2 comments
#922 - VideoMAE Model Enabling and Examples
Pull Request -
State: closed - Opened by pi314ever 5 months ago
- 7 comments
Labels: run-test
#920 - Upgrade SD output image verification with CLIP score
Pull Request -
State: closed - Opened by MohitIntel 5 months ago
- 1 comment
Labels: run-test
#919 - Update Wav2Vec2ModelTest::test_initialization
Pull Request -
State: closed - Opened by malkomes 5 months ago
- 1 comment
Labels: run-test
#918 - add fp8 related changes to mistral for text-generation
Pull Request -
State: closed - Opened by skaulintel 5 months ago
- 1 comment
Labels: run-test, synapse1.16
#917 - SW-181597 [NIGHTLY][torch.compile] Albert-XXL torch.compile full training sporadically fails with acc ~5%Test albertxxl acc
Pull Request -
State: closed - Opened by hsubramony 5 months ago
#915 - add example of multitask_prompt/poly tuning
Pull Request -
State: open - Opened by sywangyi 5 months ago
- 2 comments
Labels: review wip
#912 - Added GPT-J FP8 support
Pull Request -
State: closed - Opened by kplau1128 5 months ago
- 12 comments
#909 - Add mixtral for meta device load due to mixtral-8x22b model size
Pull Request -
State: closed - Opened by libinta 5 months ago
- 1 comment
Labels: run-test, synapse 1.16_dependency
#908 - fix perf drop in flan-t5 summarization
Pull Request -
State: closed - Opened by MohitIntel 5 months ago
- 2 comments
Labels: run-test
#907 - Integrate Transformer Engine and FP8 support
Pull Request -
State: closed - Opened by libinta 5 months ago
#906 - Set ds config "reduce scatter" to false
Pull Request -
State: closed - Opened by vineethanandh 6 months ago
- 1 comment
#905 - sasarkar/Llama3 test and readme changes
Pull Request -
State: closed - Opened by ssarkar2 6 months ago
- 1 comment
#904 - Disable wav2vec2 symbolic tracing test
Pull Request -
State: open - Opened by tjs-intel 6 months ago
#903 - Support mixtral long sequence 32k with bs 4
Pull Request -
State: closed - Opened by jychen-habana 6 months ago
- 5 comments
Labels: run-test, synapse1.16
#902 - Optimization for phi series models: support fp8 kv cache and reuse kv cache
Pull Request -
State: closed - Opened by yuwenzho 6 months ago
- 7 comments
Labels: run-test
#901 - Support mixtral long sequence 32k with bs 2
Pull Request -
State: closed - Opened by jychen-habana 6 months ago
- 7 comments
#900 - Update based on legal requirements.
Pull Request -
State: closed - Opened by libinta 6 months ago
- 1 comment
Labels: run-test
#899 - get iteration times through generate()
Pull Request -
State: closed - Opened by hsubramony 6 months ago
- 1 comment
Labels: run-test
#898 - Support mixtral kvcache reuse and remove kv_cache_fp8
Pull Request -
State: closed - Opened by jychen-habana 6 months ago
- 5 comments
Labels: run-test
#897 - Add nograd() to prevent backward backend
Pull Request -
State: open - Opened by astachowiczhabana 6 months ago
#896 - Add Support for MT5 model on Gaudi2
Pull Request -
State: open - Opened by Gaurav7888 6 months ago
- 16 comments
Labels: review wip
#895 - fix falcon 180b oom issue if peft > 0.6.2
Pull Request -
State: closed - Opened by sywangyi 6 months ago
- 1 comment
#894 - Add SDXL in README
Pull Request -
State: closed - Opened by regisss 6 months ago
- 1 comment
#893 - Fix SFT, DPO CI on Gaudi1
Pull Request -
State: closed - Opened by regisss 6 months ago
- 1 comment
#892 - Add linear and dynamic RoPE to Mistral and Mixtral
Pull Request -
State: open - Opened by regisss 6 months ago
- 1 comment
#891 - Add the MC example
Pull Request -
State: closed - Opened by yuanwu2017 6 months ago
- 14 comments
Labels: run-test
#890 - [Bug] fp8 measure failed on Llama-2-7b-hf
Issue -
State: closed - Opened by ClarkChin08 6 months ago
- 3 comments
Labels: bug
#889 - add trl example for mistral/mixtral.
Pull Request -
State: closed - Opened by lkk12014402 6 months ago
- 5 comments
#888 - Fixed "reuse_cache" Bug
Pull Request -
State: closed - Opened by Danielohayon 6 months ago
- 1 comment
Labels: run-test
#887 - text generation: added langchain pipeline script
Pull Request -
State: open - Opened by mgonchar 6 months ago
#886 - Enable HQT matmul fp8 quantization for Mistral
Pull Request -
State: closed - Opened by jiminha 6 months ago
- 1 comment
#885 - Dreambooth finetuning support
Pull Request -
State: open - Opened by srinarayan-srikanthan 6 months ago
- 2 comments
#884 - Update text-generation readme with torch.compile info.
Pull Request -
State: open - Opened by libinta 6 months ago
- 1 comment
#883 - Move pytest dependency
Pull Request -
State: closed - Opened by regisss 6 months ago
- 1 comment
#882 - Encapsulate FSDPA in GaudiLlamaAttention
Pull Request -
State: closed - Opened by dudilester 6 months ago
- 2 comments
Labels: synapse 1.16_dependency
#881 - diffuser dreambooth full/lora/lokr/loha/oft finetune, dreambooth XL lora finetune
Pull Request -
State: open - Opened by sywangyi 6 months ago
- 3 comments
#880 - Sdxl lora
Pull Request -
State: closed - Opened by atakaha 6 months ago
- 3 comments
#879 - Enabled additional models in CI
Pull Request -
State: closed - Opened by MohitIntel 6 months ago
- 3 comments
Labels: run-test
#878 - Added an example code to demonstrate how to use deterministic image generation
Pull Request -
State: closed - Opened by cfgfung 6 months ago
- 9 comments
Labels: run-test
#877 - Remove deprecated AOT_HPU_TRAINING_BACKEND (#138)
Pull Request -
State: closed - Opened by astachowiczhabana 6 months ago
- 2 comments
Labels: run-test
#876 - Test Habana's implementation of PT2E calibration and quantization feature using model Llama2
Pull Request -
State: closed - Opened by kausikmaiti 6 months ago
- 1 comment
#875 - Add mark_step for llama inference
Pull Request -
State: closed - Opened by libinta 6 months ago
- 2 comments
#874 - Skaulintel/consolidate kvcache
Pull Request -
State: closed - Opened by skaulintel 6 months ago
#873 - add first token prints
Pull Request -
State: closed - Opened by hsubramony 6 months ago
- 1 comment
Labels: run-test
#872 - After accelerate v0.28.0, the `split_batches` of class Accelerator became a unsettable property method
Issue -
State: open - Opened by eraser00 6 months ago
- 2 comments
#871 - Llama Guard for text classification example
Pull Request -
State: closed - Opened by dsmertin 6 months ago
- 5 comments
Labels: run-test
#870 - add `--trust_remote_code` for text generation examples
Pull Request -
State: closed - Opened by yangulei 6 months ago
- 3 comments
Labels: run-test
#869 - Add the Stable diffusion inpaint support
Pull Request -
State: closed - Opened by yuanwu2017 6 months ago
- 19 comments
Labels: run-test
#868 - Add textual inversion XL for Gaudi
Pull Request -
State: open - Opened by dsocek 6 months ago
- 9 comments
Labels: review wip
#867 - Added scheduler selection for SDXL fine-tuning
Pull Request -
State: closed - Opened by kplau1128 6 months ago
- 5 comments
#866 - Added Llama-2-7b fp8 text-generation test cases
Pull Request -
State: open - Opened by yeonsily 6 months ago
- 2 comments
#865 - Update to transformers 4.39.1+
Issue -
State: open - Opened by vidyasiv 6 months ago
- 2 comments
#864 - Jwang/fix convert into dtypes
Pull Request -
State: closed - Opened by IT-Forrest 6 months ago
#863 - Fp8 merge fix
Pull Request -
State: closed - Opened by libinta 6 months ago
- 1 comment
#862 - Enable Flash Attention in recompute and causal modes (#21)
Pull Request -
State: closed - Opened by wszczurekhabana 6 months ago
- 2 comments
Labels: run-test