cmedina/agenx-lora-training
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: load bf16 model, quantize with BnB on CPU, move to GPU

Browse Source
This commit is contained in:
Christian Medina
2026-07-02 19:55:01 -04:00
parent a0f4f644b0
commit 3c31e4dfb0
2 changed files with 37 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
training/configs/ornith-35b-lora.yaml
View File

@@ -1,7 +1,7 @@
# LoRA Training Configuration for Ornith-1.0-35B
# Dataset: cyron_summary_lora_dataset (20k examples)
base_model: /data/models/Ornith-1.0-35B-4bit
base_model: /data/models/Ornith-1.0-35B
model_type: Qwen3_5MoeForCausalLM
tokenizer_type: AutoTokenizer