Add eval_batch_size for evaluation#12
Merged
Merged
Conversation
8 tasks
djsaunde
pushed a commit
that referenced
this pull request
Dec 17, 2024
Add eval_batch_size for evaluation
OnePunchMonk
added a commit
to OnePunchMonk/axolotl
that referenced
this pull request
Mar 22, 2026
- #1: Rename original_ffn -> base_ffn (fixes AttributeError in tests) - axolotl-ai-cloud#2: Fix device mismatch in collect_mixlora_aux_loss zero tensor - axolotl-ai-cloud#3: Remove > 0 guard on aux loss (always add and log) - axolotl-ai-cloud#4: Fix or-fallback bug for falsy numeric config values - axolotl-ai-cloud#5: Remove unnecessary .clone() in MixLoraFFN.forward - axolotl-ai-cloud#6: Fix MixLoraExpert.forward return type annotation - axolotl-ai-cloud#7: Update comments for down_proj delta computation - axolotl-ai-cloud#8: Add inference guard (eval mode) for MixLoRA in adapter loader - axolotl-ai-cloud#9: Replace @pytest.mark.skip with @pytest.mark.slow - axolotl-ai-cloud#10: Add ge/gt field constraints and top_k <= num_experts validator - axolotl-ai-cloud#11: Block flash_attn_fuse_qkv with mixlora - axolotl-ai-cloud#12: Add mixlora_state_dict/load_mixlora_state_dict for checkpointing - axolotl-ai-cloud#13: Fix typo intermmediate -> intermediate
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem:
I would OOM during the evaluation stage despite setting
micro_batch_sizeto be low. Turns out that batch size for eval using a different parameter.Proposal:
New config for eval stage for lower VRAM GPUs.
Edit:
Since it's now a somewhat required argument, I'm not sure whether we should