Jwang/fix convert into dtypes#864
Closed
IT-Forrest wants to merge 83 commits into
Closed
Conversation
Co-authored-by: Sayantan Sarkar <sasarkar@habana.ai> Co-authored-by: Libin Tang <litang@habana.ai> Co-authored-by: Jimin Ha <jha@habana.ai> Co-authored-by: Yeonsil Yoon <yyoon@habana.ai> Co-authored-by: Sayantan Sarkar <supersarkar@gmail.com>
* Expose Llama Fused OPs control from run_lora_clm.py * Update as per review comments
* enable internal kv bucket in llama * initialize bucket_internal for CI * make bucket_internal more clear * further perf optim while max length is not multiple of bucket size
* [SW-173358] add first token prints * [SW-173358] rename x to outputs * [SW-173358] make style
* Enable Flash Attention in recompute and causal modes * Add flash_attention_causal_mask to generation utils * Propagate Flash Attention causal_mask to finetuning example * Modify README example and provide additional description * Add flash_attention_causal_mask to FT README
* enable loading falcon-180b ckpt in .safetensors format * Address comments borrowing transformer's way of reading ckpt file * address comments
Co-authored-by: Sun Choi <schoi@habana.ai>
* enable loading falcon-180b ckpt in .safetensors format * Address comments borrowing transformer's way of reading ckpt file * address comments * Update ckpt loading PR#15 reads a set of ckpt file names from the index json file. When OH downloads files from the hub instead of loading from a cache dir, get_repo_root() skips downloading the index json file. Thus the PR#15 fails to load file names. This PR scans the path and returns a list of names that matches the pattern * import modeling_utils from transformers
* enable Falcon FP8 inference * added example command in readme, code cleanup * resolve issues in finetuning * enable non reuse cache flow for fp8 * revert non reuse_cache flow for training due to perf drop --------- Co-authored-by: Local Lab User <labuser@habana-labs.com>
…=1 (#96) * Added additionla check to run with distributed enabled and world_size = 1 * Reduce the number of graph splits to avoid memory allocation error for 1x LLAMA1_7b_ft --------- Co-authored-by: Kalyan <kkumar@habana.ai>
Update utils.py
Co-authored-by: Kalyan <kkumar@habana.ai>
* enable Falcon FP8 inference * added example command in readme, code cleanup * resolve issues in finetuning * enable non reuse cache flow for fp8 * revert non reuse_cache flow for training due to perf drop * add falcon180B FP8 test * fix error * fix run_lm_eval.py to save --reuse_cache * fix Falcon view+inplace error --------- Co-authored-by: Local Lab User <labuser@habana-labs.com>
…at matches its scale method (#92)
* Done to allow quantization using HQT * Added use_flash_attention and flash_attention_recompute to run_lm_eval
IT-Forrest
requested review from
ZhaiFeiyue,
bhargaveede,
hlahkar,
libinta,
mandy-li,
ssarkar2 and
vivekgoe
as code owners
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This is the cherry-pick of the Optimum-habana commit #769. The fix consists in extracting the dtype of the first logit tensor and set it as the target dtype. We used to extract the dtypes of all logit tensors, leading to nested tuples that led to this issue. There is no reason for logits to have different dtypes, so we can simplify the dtype extraction logic.
Fixes # (754)
Before submitting