Merged
Conversation
rolandtannous
added a commit
to rolandtannous/unsloth
that referenced
this pull request
Oct 9, 2025
danielhanchen
pushed a commit
that referenced
this pull request
Nov 20, 2025
* make loading gpt-oss-BF16 faster. Linked to unsloth-zoo PR #314 * fix model loading and clean merged model directory * revert default quant * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * revert mapper.py --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
danielhanchen
added a commit
that referenced
this pull request
Nov 25, 2025
* Enable FP8 + RL training for bf16 models (#3440) * Enable FP8 + RL training for bf16 models **Summary:** Enable FP8 + RL training using TorchAO for 1.33x faster training and 42% less model memory usage: - We quantize the frozen LoRA weights into fp8 and keep the LoRA adapters in bf16 - We leverage TorchAO's `Float8Tensor`, which calls into fbgemm's fp8 x fp8 rowwise matmul kernel - For now, we need to do an offline quantization first, because vllm doesn't support on-the-fly quantization for torchao yet (this is in progress: vllm-project/vllm#26327) **Example usage:** ``` model, tokenizer = FastLanguageModel.from_pretrained( model_name = "unsloth/Qwen3-8B-Base", max_seq_length = 2048, load_in_4bit = False, fast_inference = True, max_lora_rank = 32, load_in_fp8 = True, # set this to True ) \# the rest is the same as before model = FastLanguageModel.get_peft_model(...) ``` **Initial results:** ``` \# fp8 {'train_runtime': 1725.4337, 'train_samples_per_second': 0.232, 'train_steps_per_second': 0.058, 'train_loss': 0.00015715716748673002, 'epoch': 0.01} \# bf16 {'train_runtime': 2297.8145, 'train_samples_per_second': 0.174, 'train_steps_per_second': 0.044, 'train_loss': 0.00016081033063528594, 'epoch': 0.01} ``` <img width="1199" height="448" alt="Screenshot 2025-11-11 at 4 10 50 PM" src="https://github.com/user-attachments/assets/b6304afd-89e9-42b1-8064-775807e17b23" /> Test script: https://gist.github.com/andrewor14/5b85119fae46845d07b608d420907423 **Requires:** - pytorch/ao#3158 (torchao nightly or 0.15.0+) - unslothai/unsloth-zoo#351 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update utils.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * _get_inference_mode_context_manager * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * Update utils.py * Update utils.py * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Daniel Han <danielhanchen@gmail.com> * Update __init__.py * Fix/save torchao model loading logic (#3621) * make loading gpt-oss-BF16 faster. Linked to unsloth-zoo PR #314 * fix model loading and clean merged model directory * revert default quant * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci * revert mapper.py --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Update loader_utils.py * Update loader_utils.py * Add 128x128 PerBlock FP8 + RL (#3629) * Add 128x128 PerBlock FP8 + RL **Summary:** Following #3440, this PR extends torchao FP8 + RL support to also handle 128x128 PerBlock granularity (in addition to PerRow). **Example usage:** ``` model, tokenizer = FastLanguageModel.from_pretrained( model_name = "unsloth/Qwen3-8B-Base", max_seq_length = 2048, load_in_4bit = False, fast_inference = True, max_lora_rank = 32, load_in_fp8 = "block", # or "row" or True ) ``` **Initial results:** TBD **Note:** - Requires pytorch/ao#3370 * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> * Version * Update vision.py * Update rl.py * Add torch 2.9.1 * Fix auto installer * Update fp8.py * Float8 * Update fp8.py * Update mapper.py * Update mapper.py * Update loader_utils.py * Update loader.py * Update fp8.py * Versioning * [pre-commit.ci] auto fixes from pre-commit.com hooks for more information, see https://pre-commit.ci --------- Co-authored-by: andrewor14 <andrewor14@gmail.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Roland Tannous <115670425+rolandtannous@users.noreply.github.com>
rolandtannous
added a commit
to rolandtannous/unsloth
that referenced
this pull request
Mar 11, 2026
…rsion-error-handling-local Fix VLM training abort on URL-based dataset conversion failure
rolandtannous
added a commit
that referenced
this pull request
Mar 12, 2026
…or-handling-local Fix VLM training abort on URL-based dataset conversion failure
Stanley00
pushed a commit
to stanley-fork/unsloth
that referenced
this pull request
Mar 12, 2026
…rsion-error-handling-local Fix VLM training abort on URL-based dataset conversion failure
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
1 participant
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.
No description provided.