huggingface/optimum-habana

Issues

• Error when running chatglm3_6b: NotImplementedError: Unknown device for graph fuser

  #1477 opened 11 days ago by BaideBear
  1

• Pretrain with LLama Model - num_samples=0 Error

  #1396 opened 16 days ago by saisuryateja1436
  9

• RuntimeError: [Rank:0] FATAL ERROR :: MODULE:PT_DEVMEM Allocation failed for size::234881024 (224)MB #2

  #1469 opened 21 days ago by James-Lu-none
  2

• Error when running llama2_fine_tuning_inference & Intel_Gaudi_Fine_Tuning examples

  #1467 opened 23 days ago by epage480
  1

• Could not find a version that satisfies the requirement networkx==3.3 in Python3.9

  #1458 opened a month ago by James-Lu-none
  1

• Qwen1.5-14B finetune error

  #1336 opened 2 months ago by Zjq9409
  2

• Meta-Llama-3 model text-generation example output is unexpected on 2 nodes

  #1451 opened a month ago by aslanxie
  4

• meta-llama/Llama-3.1-8B-Instruct generate output shows unexpected padding

  #1378 opened a month ago by aslanxie
  3

• 8 cards inference with input length 50k with device critical error

  #1035 opened 6 months ago by ClarkChin08
  1

• Performance for summarization task on BART is low after latest Transformer 4.40 upgrade

  #1144 opened 4 months ago by astachowiczhabana
  1

• Heavy IO in multi-node example

  #1152 opened 4 months ago by rofinn
  1

• Tokenizer error: eos_token_id not found error - Incorrect assignment of variables

  #1326 opened a month ago by premmotgi
  4

• Llama 3.1 Support -- Rope_scaling issue

  #1154 opened a month ago by AzeezIsh
  6

• Not able to get good performance for diffusion models when doing single image inference with batch size 1

  #1195 opened 4 months ago by basantaxpatra
  1

• Finetuning Llama3.1 70B Model OOM error with LoRa technique

  #1327 opened 2 months ago by premmotgi
  1

• ValidateSyncInputTensors tensor_data is empty

  #1241 opened 3 months ago by xinsu626
  1

• Hard-coded llama `pad_token_id` in `examples/text-generation/utils.py`

  #1170 opened a month ago by muhark
  1

• quantization FP8 error

  #1438 opened a month ago by aitss2017
  1

• bigscience / bloomz-7b1 finetune error

  #1426 opened a month ago by 11989890
  2

• Transformers 4.45.0 incompatible with Optimum Habana 1.14.0.dev0

  #1417 opened a month ago by Sanatan-Shrivastava
  7

• transformers_future: contrastive search failing with Incompatible input shapes, broadcast not possible.

  #1385 opened a month ago by vidyasiv
  2

• Some non-default generation parameters are set in the model config

  #1351 opened a month ago by vidyasiv
  6

• Llama-2-70B FP8 quantization trust_remote_code=True not passed in llm_eval

  #1321 opened a month ago by aalbersk
  1

• RuntimeError: shape '[-1, 0]' is invalid for input of size 2510226432

  #1306 opened a month ago by xinsu626
  1

• Data parallelism or tensor parallelism? How can i know that and is there a chance i can shift in between these too?

  #1242 opened a month ago by venkycreator
  1

• Info needed about stable diffusion 3 support

  #1127 opened a month ago by dkiran1
  2

• does the optimum-habana support sdxl controlnet image to image pipeline?

  #1103 opened a month ago by basantaxpatra
  2

• Is there example of FP8 train LLM, pre-train or fine-tune

  #1073 opened a month ago by harborn
  1

• AttributeError: module 'transformers.generation.stopping_criteria' has no attribute 'MaxNewTokensCriteria'

  #1372 opened 2 months ago by hemanthkotaprolu
  2

• transformer 4.45 -> causing image_to_text failures for gaudi attribute error

  #1357 opened 2 months ago by jiminha
  7

• Dataset v3.0.0 deprecates tasks and cause CI failures

  #1341 opened 2 months ago by vidyasiv
  4

• Flash attention not supported in run_clm.py

  #1318 opened 2 months ago by aitss2017
  5

• RuntimeError: shape '[-1, 0]' is invalid for input of size 134152192 for Mistral-7B finetune

  #1311 opened 2 months ago by aitss2017
  6

• runwayml/stable-diffusion-v1-5 no longer exists

  #1305 opened 3 months ago by vidyasiv
  3

• CodeGen inference error "synNodeCreateWithId failed for node: batch_gemm with synStatus 26"

  #1314 opened 3 months ago by caijimin
  4

• stable-diffusion-2-1-base BF16 on Gaudi2D get RuntimeError: synNodeCreateWithId failed for node: spatial_convolution with synStatus 26 [Generic failure].

  #1293 opened 3 months ago by KiwiHana
  3

• CLIP contrastive image-text inference can't run on hpu with gaudi-docker/1.17.0

  #1216 opened 3 months ago by caijimin
  12

• AWQ is not working

  #1240 opened 3 months ago by endomorphosis
  4

• Qwen2-72B inference on 8x Gaudi2 gets OOM issue due to missing meta-device support on model loading

  #1112 opened 3 months ago by LeoZhao-Intel
  6

• Quantization failed

  #1237 opened 3 months ago by endomorphosis
  6

• CLIP contrastive image-text inference error

  #1179 opened 4 months ago by caijimin
  1

• Batch size beyond 16 is throwing an error

  #1177 opened 4 months ago by venkycreator
  2

• llava inference works incorrectly if adapt_transformers_to_gaudi called after transformers import

  #1176 opened 4 months ago by mattolson93
  2

• The name of the "is_greedy_or_beam_and_bucket" variable does not seem to match the logic

  #1166 opened 4 months ago by Jing1Ling
  2

• TypeError: GaudiPhiForCausalLM.forward() got an unexpected keyword argument 'reuse_cache'

  #1036 opened 4 months ago by eduand-alvarez
  3

• Support for auto-language-detection Whisper inference on HPU

  #1049 opened 6 months ago by Spycsh
  1

• Unable to run protein-folding example with the latest release: optimum-habana-1.11.1

  #1014 opened 5 months ago by mgujral
  25

• --report_to tensorboard not working on multiple HPUs

  #1021 opened 5 months ago by 12010486
  1

• Qwen1.5-32B hits GC compilation failed in some cases.

  #1012 opened 5 months ago by LeoZhao-Intel
  4

• Add support on Whisper inference on HPU

  #1047 opened 6 months ago by Spycsh
  0