phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-22 18:46:16 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: Update SFTConfig parameter

Browse source 
- Change max_seq_length to max_length in SFTConfig constructor
- TRL deprecated max_seq_length in Feb 2024 and removed it in v0.20.0
- Reference: https://github.com/huggingface/trl/pull/2895

This resolves the SFT training failure in CI tests
This commit is contained in:
Nehanth 2025-07-29 18:00:50 +00:00
parent 870a37ff4b
commit 3c510750aa
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llama_stack/providers/inline/post_training/huggingface/recipes/finetune_single_device.py
View file

@ -469,7 +469,7 @@ class HFFinetuningSingleDevice:
use_cpu=True if device.type == "cpu" and not torch.backends.mps.is_available() else False, use_cpu=True if device.type == "cpu" and not torch.backends.mps.is_available() else False,
save_strategy=save_strategy, save_strategy=save_strategy,
report_to="none", report_to="none",
max_seq_length=provider_config.max_seq_length, max_length=provider_config.max_seq_length,
gradient_accumulation_steps=config.gradient_accumulation_steps, gradient_accumulation_steps=config.gradient_accumulation_steps,
gradient_checkpointing=provider_config.gradient_checkpointing, gradient_checkpointing=provider_config.gradient_checkpointing,
learning_rate=lr, learning_rate=lr,