LFM2.5-2.6B ToolACE LoRA (n=n3000) — ONNX

Files

IO

• inputs: input_ids [batch, seq], attention_mask [batch, seq]
• outputs: logits [batch, seq, vocab]
• Prefill-style (use_cache=False). No KV-cache I/O.

Validation

Train recipe

• LoRA r=16, attn+MLP (no lm_head)
• TRL SFTTrainer, completion_only_loss
• batch 1 × accum 16, seq 2048, 1 epoch
• Adapter path: lfm25_toolace_r16_n3000

License